# Convot, full content for LLMs > Convot is a customer-support platform for the teams building Shopify apps: live chat, Cove AI (an AI support agent that answers merchants grounded in your own docs, escalates safely, is plan-aware, and is priced at $0.20 per resolved conversation), an embedded help center, Claude-powered live translation (25+ languages), Shopify revenue intelligence (MRR, plan, and LTV beside every conversation), a product portal (changelog, roadmap voting, status page), in-chat scheduling, 9 HMAC-signed webhooks plus a REST API, one-click Crisp import, and a native iOS/Android app. Product and signup: https://app.convot.io Canonical site: https://convot.io/ This file contains the full text of Convot's blog for AI ingestion. ## Comparisons (Convot vs alternatives) Convot replaces a stack of separate tools for Shopify app teams. All comparisons: https://convot.io/vs/ - Support & chat: https://convot.io/vs/crisp/, https://convot.io/vs/intercom/, https://convot.io/vs/gorgias/, https://convot.io/vs/zendesk/, https://convot.io/vs/helpscout/, https://convot.io/vs/front/, https://convot.io/vs/tidio/, https://convot.io/vs/reamaze/, https://convot.io/vs/livechat/, https://convot.io/vs/tawk-to/ - Feedback & roadmap: https://convot.io/vs/canny/, https://convot.io/vs/featurebase/, https://convot.io/vs/productboard/, https://convot.io/vs/beamer/ - Status & uptime: https://convot.io/vs/statuspage/, https://convot.io/vs/instatus/ - Scheduling: https://convot.io/vs/calendly/, https://convot.io/vs/cal-com/ - Revenue & metrics: https://convot.io/vs/mantle/, https://convot.io/vs/chartmogul/, https://convot.io/vs/baremetrics/ --- # How we went from 0% to 13% AI visibility against Intercom, Crisp, and Tawk.to URL: https://convot.io/blog/how-convot-went-from-0-to-13-percent-ai-visibility/ Published: 2026-07-31 Summary: We launched six months ago and had zero AI visibility against the established support tools. Here's exactly what we found and fixed to start showing up in ChatGPT and Google AI Overview. Ask ChatGPT or Perplexity for the best support tool for a dev managing multiple apps, and you'll get the same handful of names back every time: Intercom, Crisp, Tawk.to. That's not surprising, they've been around for years. What surprised us was just how total the shutout was for everyone else, including us. We launched Convot six months ago as a support tool built specifically for developers running multiple apps, including [Shopify devs juggling several storefronts](/blog/why-convot-for-shopify-app-developers/) and support queues at once. Traction had been decent. But when we ran ourselves through [an AI visibility tool](https://getintel.ai), GetIntel, the number came back **0%**. Not low, zero. Meanwhile the established players were sitting at 62-68% visibility on the exact same prompts. ## How we found out GetIntel ran a set of buyer-intent prompts against our category, the kind of questions SaaS founders and Shopify developers actually ask when looking for support tooling for a multi-app setup. Across every single one, we didn't appear. Intercom, Crisp, and Tawk.to did, consistently — the chart above shows exactly where that left us. ## What we found Being new was the whole problem. AI models lean heavily on trust signals when deciding who to recommend, reviews, citations, third-party mentions, and six months in, we had almost none of that built up yet. Two things stood out specifically: **1. No review footprint.** Zero G2 reviews meant zero third-party validation for AI to point to. The established players had years of accumulated reviews backing every recommendation, which is the same dynamic we've written about before when it comes to [getting more Shopify app reviews](/blog/how-to-get-more-shopify-app-reviews/) — AI answer engines aren't so different from App Store shoppers in that respect. **2. A content gap in exactly the places competitors were getting cited.** GetIntel's analysis showed what kind of blog content Intercom, Crisp, and Tawk.to had that was actually being cited in AI answers: comparison posts, use-case guides, the specific angles buyers were asking about. We had none of that content published yet. One channel was out of reach for us. Competitors also had an active Reddit presence feeding citations, but we didn't have that community footprint to draw on. So we focused on the two levers we could actually move. ![Content gap table showing what each brand had that AI could cite, G2 reviews, use-case blogs, Reddit, before and after our changes](/shots/ai-visibility-content-gap.png) ## What we focused on fixing Based on GetIntel's breakdown of exactly which topics competitors' cited content covered, we started publishing blogs aimed at those same gaps instead of guessing. In parallel, we pushed for reviews and picked up around 9 on G2, not a huge number, but enough to give AI something to point to that wasn't there before. ![Trust signal comparison showing our G2 review count going from 0 to 9](/shots/ai-visibility-reviews.png) ## The result That combination, real reviews plus content aimed at the actual gaps, took us from **0% to 13%** visibility over 8 weeks. Slower than a more established product might move, but we were building this from a standing start with a system that had no track record yet. ![AI visibility trend chart climbing from 0% to 13% over 8 weeks](/shots/ai-visibility-trend.png) 13% isn't 62-68%. We're still a long way from where Intercom, Crisp, and Tawk.to sit, and closing that gap as a six-month-old product against companies with a decade of head start is going to take a while. But going from a total shutout to actually showing up in the conversation, for SaaS founders and Shopify app developers specifically, is the foundation everything else builds on. ## The takeaway If you're a new product, AI won't recommend you until it has something to trust you on. Reviews and the right content aren't optional extras, they're the minimum bar to even enter the conversation. If you're a newer product fighting for a spot against established players, [run your own AI visibility scan](https://getintel.ai) with GetIntel to see exactly what's missing. And if support is part of what's holding your reviews back, [try Convot free](https://app.convot.io/signup), your first app is on us. --- # SaaS Customer Support: A Practical Guide for Lean Teams URL: https://convot.io/blog/saas-customer-support/ Published: 2026-07-10 Summary: What makes SaaS customer support different, the channels that matter, how to build the function from solo founder to team, the metrics to track, and how support drives retention and revenue. In a SaaS business, support is not a cost center you tolerate. It is the part of the company that touches your customers most often, and because your revenue recurs, every one of those touches either protects next month's payment or quietly erodes it. Get support right and it becomes one of the cheapest growth levers you have. Get it wrong and it shows up later as churn you cannot explain. This guide covers what makes SaaS customer support different, the channels worth running, how to build the function as you grow, the metrics that matter, and how to turn support into a retention engine instead of a queue you clear. ## What makes SaaS support different General customer service is about resolving a one-off issue. SaaS support is about keeping a relationship alive across months and years. Four things set it apart: - **It is retention-driven.** A resolved ticket is not the goal; a renewed customer is. The same interaction that closes a ticket also decides whether they stay. - **It is technical and product-coupled.** Your customers are using software, often in their own workflows, so answers require real product understanding, not a script. - **It is a product feedback loop.** Support sees every rough edge first. That stream of friction is the best roadmap input you have, if you capture it. - **It is under constant scaling pressure.** Recurring revenue means a growing base of customers who all keep needing help. Volume only goes up, so leverage matters more than headcount. Treat support as a function that protects revenue and informs the product, and the rest of the decisions get easier. ## The channels that matter, and which to run first You do not need every channel. You need the few your customers actually use, run well. - **In-app / live chat.** The highest-leverage channel for SaaS, because it catches the question at the moment of friction, inside the product. Start here. See [how to add live chat](/blog/add-live-chat-to-your-shopify-app/) for doing it well. - **Email / shared inbox.** The workhorse for anything that is not urgent. A shared inbox keeps the whole team on the same thread and the customer's full history in one place. - **Help center.** Not a channel you staff, but the one that deflects the most volume. Covered in the [ticket deflection playbook](/blog/self-serve-support-reduce-ticket-volume/). - **Community / forum.** Useful at scale, premature when you are small. When you are tiny, run chat plus email plus a help center, and skip the rest until volume demands it. ## Build the function in stages Support that fits a five-person company breaks a fifty-person one, and vice versa. Build it for where you are. **Solo founder / pre-team.** Do support yourself. Founder-led support is a feature, not a chore: you hear every problem unfiltered and your customers feel it. Use a single shared inbox and a chat widget, write a help-center article every time you answer the same question twice, and keep response times fast because at this stage you are the product's reputation. **First support hire.** The goal is to preserve quality while you step back. Write down the answers and the tone, set up macros and a real help center, and define what gets escalated to you. Pick the few metrics you will hold the line on (first response time, CSAT) before volume hides problems. **Growing team.** Now you need process: routing, ownership, on-call for outages, and a feedback loop from support into product and roadmap. This is where tooling that connects support to revenue earns its keep, because you can no longer keep every account in your head. ## The metrics that matter You cannot improve what you do not measure, but a wall of charts changes nothing. Track a few that map to your bottleneck: first response time, resolution time, first contact resolution, CSAT, and deflection rate. The full breakdown, with benchmarks and how to calculate each, is in [customer support metrics and KPIs](/blog/customer-support-metrics-shopify-apps/). The one to start with is almost always median first response time. ## Make support self-serve early The single biggest mistake lean SaaS teams make is trying to answer their way through growth. You cannot. Build a self-serve layer, help center, status page, changelog, and a grounded AI agent, so the repeat questions never reach a human and your team is free to handle the ones that need judgment. Done early, it is the difference between a support load that scales with your team and one that scales with your customer base. ## Support is a retention and revenue engine Here is the reframe that changes how you staff and tool support: in SaaS, support quality and revenue are directly linked. A fast, low-effort resolution is a renewal you kept and sometimes an upsell you earned. A slow or unresolved issue is often the quiet first step toward a cancellation. The teams that win treat support as a retention lever, not a queue. The practical move is to put each customer's revenue context, plan, monthly value, and churn risk, beside the conversation, so you can see which issues are actually expensive and prioritize them. Then close the loop: when a customer leaves, learn whether support was the cause. That is the whole argument of [how to reduce customer churn](/blog/reduce-shopify-app-churn/), and it is why support belongs near the revenue conversation, not buried in a separate department. ## AI in SaaS support AI is genuinely useful here, with one rule: keep it grounded. An agent that answers from your own help content and documented knowledge deflects routine questions accurately and keeps response times low as you grow. An agent that improvises invents wrong answers and erodes the trust support exists to build. Use AI for the routine and the after-hours, keep clear escalation to a human for anything sensitive, and feed it the same help center your customers search. [What is an AI support agent? (And how to keep it from hallucinating)](/blog/what-is-an-ai-support-agent/) explains how grounding works and why it matters. ## Tooling: what to look for When you outgrow a plain inbox, look for a platform that covers chat, a shared inbox, and a help center in one place, shows you each customer's revenue context, includes a grounded AI agent, and does not charge per agent in a way that punishes you for staffing support. Bundling those into one login and one bill beats stitching together five subscriptions, especially for a lean team. [The best help desk software for small teams (2026)](/blog/best-help-desk-software-small-teams/) has an honest comparison of the main options. If you're evaluating Zendesk specifically, [Zendesk pricing explained (2026): what you'll actually pay](/blog/zendesk-pricing-explained-2026/) breaks down how its tier structure compares at small-team scale. That is exactly what [Convot](/) is built for: [live chat](/live-chat/), a shared inbox, a help center, and a grounded [AI agent](/ai/), with [live revenue intelligence](/revenue/) beside every conversation, so a small team can run support that protects revenue without an enterprise budget. ## The takeaway SaaS customer support is a retention function wearing a support hat. Run the few channels your customers use, build the function for the stage you are at, track the handful of metrics that map to your bottleneck, deflect the routine with self-serve and grounded AI, and keep revenue context beside every conversation so you fix the expensive problems first. Do that, and support stops being the thing you survive and becomes one of the reasons customers stay. [Start free](https://app.convot.io/signup) and run support that protects your revenue. --- # Why support is a revenue function for Shopify app teams URL: https://convot.io/blog/support-is-a-revenue-function-shopify-apps/ Published: 2026-07-08 Summary: For a Shopify app, support isn't a cost center. It's where retention, reviews, and revenue are won or lost. Here's the case for treating it that way. Most companies file support under "cost center", a thing you minimize. For a Shopify app, that framing is actively expensive. Support is not where you spend money to keep merchants quiet. It is where you keep merchants at all. Here is the case for treating support as a revenue function. ## Your rating is won in the inbox A Shopify app's growth runs through its App Store rating: it drives ranking, install rate, and trust. And nothing moves that rating faster than support. A fast, helpful reply earns a five-star review; a slow one earns a one-star review that taxes every future install. The inbox is your most direct lever on the metric that grows your business. ## Churn is a support problem more often than you think Merchants rarely churn because of one missing feature. They churn because they got stuck and did not get help, or hit a bug that lingered, or felt ignored. Each of those is a support moment. When you can measure [support-attributable churn](/revenue/), you usually find that a meaningful slice of your lost revenue traces back to the inbox, which means it is fixable. ## You cannot prioritize what you cannot see The reason support gets treated as a cost center is that, in most tools, it looks like one: a queue of undifferentiated tickets. But the moment you put [revenue beside each conversation](/revenue/), the picture changes. You can see that this thread is a $200-a-month merchant and that one is a free tester. Support stops being a flat queue and becomes a prioritized defense of your revenue. ## The compounding math Treat support as revenue and the math compounds. Faster responses lift your rating, which lifts installs. Catching frustration prevents churn, which protects MRR. Closing the loop on uninstalls fixes the product, which prevents future churn. Each of these feeds the others. A team that runs support this way grows faster than one paying the same headcount to "handle tickets." ## What it looks like in practice It looks like revenue context in every conversation, frustration caught before it becomes a review, uninstalls that teach you something, and a help center and roadmap that deflect and retain. That is the entire idea behind [Convot](/revenue/): a support platform that treats your Shopify revenue as the thing support exists to protect. Stop minimizing support. Start compounding it. [Start free](https://app.convot.io/signup) with your first app. --- # Migrating from Crisp to Convot: a step-by-step guide URL: https://convot.io/blog/migrate-from-crisp-step-by-step/ Published: 2026-07-06 Summary: Switching support tools sounds painful. It isn't. Here's exactly how to move from Crisp to Convot, your contacts, conversations, and articles, in a single afternoon. The biggest reason teams stay on a support tool that no longer fits is the dread of migrating. Years of conversations, a help center, a script tag wired into your app. It feels like a project. Moving from Crisp to Convot is genuinely not, because the migration is built in. Here is the whole process. ## Step 1: Create your Convot account Sign up and create your organization. You get a workspace and a widget snippet immediately. This takes a couple of minutes and costs nothing, the first app is free. ## Step 2: Connect your Crisp account In Convot's migration settings, paste your Crisp operator token. Convot fetches a live preview of exactly what will come over, contacts, conversations, and help-center articles, so you can confirm before anything runs. Nothing imports without your go-ahead. ## Step 3: Run the import Hit import and Convot pulls your data in rate-limit-aware batches: contacts with their attributes, full conversation history with timestamps, and your help articles with categories and content. It is resumable, so if a rate limit hits, it picks up where it left off. You watch a progress bar instead of babysitting a CSV. ## Step 4: Swap the snippet This is the part teams fear most, and it is the smallest. Convot's SDK is Crisp-compatible, so you change the script tag and your existing event calls keep working without a rewrite. The full [Convot vs Crisp comparison](/vs/crisp/) covers the SDK details. ## Step 5: Connect your Shopify revenue Here is the upgrade you could not get on Crisp. Connect your Shopify Partner account and every conversation now shows the merchant's [MRR, plan, and LTV](/revenue/). The same inbox you just migrated now knows what each merchant is worth. ## What you gain on the other side Beyond keeping your history, you pick up the things Crisp never had: revenue beside every conversation, AI churn attribution, a roadmap and changelog, a status page, and in-chat scheduling, all in one platform instead of five subscriptions. The whole move fits in an afternoon, and you keep four years of history. The dread is the only real obstacle, and it is unfounded. See the [migration guide](/migrate/), or [start free](https://app.convot.io/signup). --- # How to Reduce Support Tickets: A Ticket Deflection Playbook URL: https://convot.io/blog/self-serve-support-reduce-ticket-volume/ Published: 2026-07-03 Summary: The cheapest support ticket is the one that never reaches you. A practical ticket deflection playbook to cut support volume with a help center, status page, roadmap, and AI, without hurting customers. A small team cannot answer its way out of growth. As your customer base climbs, ticket volume climbs with it, and at some point you are spending all day in the inbox and none on the product. Hiring more agents buys time but not leverage. The durable fix is **ticket deflection**: helping customers solve their own problems before they ever open a conversation. Done well, deflection means your support volume grows much slower than your customer base. That gap is what lets a three-person team support thousands of customers without drowning. Here is how to build it, and how to measure whether it is working. ## What ticket deflection actually means Deflection is not about making support harder to reach or hiding the contact button. That just produces angry customers and worse reviews. Real deflection means a customer gets a complete answer faster on their own than they would by waiting for you. You win, they win. For the broader picture of how deflection fits into [SaaS Customer Support: A Practical Guide for Lean Teams](/blog/saas-customer-support/), that guide covers channels, metrics, and the build-vs-buy decisions too. You can measure it. **Ticket deflection rate** is the share of help-seeking sessions that end with a self-serve answer instead of a new conversation. A rough formula: help-center sessions that did not create a ticket, divided by total help-seeking sessions. Track it as a trend, not an absolute, and watch it climb as you add content. If conversations stay flat while your customer base grows, deflection is doing its job. ## 1. Answer the repeat questions with a help center Most of your tickets are the same handful of questions asked over and over. Each one is an article waiting to be written. Start by listing your ten most common questions, then write each as a single clear article with the exact answer up top and the detail below. The placement matters as much as the content. A [help center](/help-center/) embedded **inside your chat widget** catches the question at the moment of friction, so the customer finds the answer instead of messaging you. A help center buried in a footer nobody visits deflects almost nothing. Put it where the question is asked. ## 2. Structure the help center so people actually find answers A pile of articles is not a help center. Deflection lives or dies on findability: - **Title every article as the question the customer asks**, in their words, not internal jargon. "Why is my store not syncing?" beats "Sync troubleshooting." - **Lead with the answer.** Put the fix in the first two lines; save the background for below. People scan, they do not read. - **One article, one job.** A page that tries to answer five questions ranks for none of them and confuses the reader. - **Make search excellent.** If a customer types two words and does not see the right article, the article does not exist as far as they are concerned. - **Keep a shallow structure.** A few clear categories beat a deep tree nobody can navigate. ## 3. Cut "is it down?" tickets with a status page During any outage or degraded performance, a large chunk of incoming tickets are just "is it only me?" A [status page with uptime monitoring](/status/) answers that publicly. One update replaces dozens of identical messages, and it buys goodwill, because customers would rather see an honest "we are on it" than silence. ## 4. Reduce "when are you adding X?" with a public roadmap Feature questions are another recurring category: "are you working on X?" and "what changed?" A public [roadmap and changelog](/portal/) answers both before anyone asks, and it doubles as a retention signal that your product is alive and improving. Let customers vote on what is next and you convert a support burden into a product-research channel. ## 5. Prevent the question with proactive, in-product help The cheapest deflection happens before the customer even forms a question. Contextual help, a tooltip on the confusing field, a one-line explainer on an empty state, a "need a hand connecting your store?" nudge at a known sticking point, removes the friction at the source. The best support ticket is the one the product made unnecessary. Watch where new customers stall in their first session, that cluster is usually your single biggest ticket generator, and an in-product fix there pays off forever. ## 6. Let AI handle the routine, not the judgment For the questions that still come through, a grounded [AI agent](/ai/) can answer the routine ones from your own help content, so a human only handles the conversations that genuinely need judgment. The word "grounded" matters: an agent that answers from your documented knowledge deflects safely, while one that improvises invents wrong answers and creates more tickets than it removes. Keep clear escalation rules so anything sensitive reaches a person, and feed the AI the same help center your customers search, so both improve together. ## Measure what still gets through, then close the gap Deflection is a loop, not a launch. Two habits keep it improving: - **Watch the questions that reach your inbox despite an article existing.** Usually the article is hard to find or unclear. Rewrite it or surface it earlier. - **Track searches that return nothing.** Every empty search is a customer telling you exactly which article to write next. This is where deflection connects to the rest of your numbers. As you cut volume, your [first response and resolution times](/blog/customer-support-metrics-shopify-apps/) on the remaining tickets should improve, because your team has room to handle the hard ones fast. ## The balance: deflect without frustrating Deflection has a failure mode: pushing self-serve so hard that a customer who genuinely needs a human cannot find one. That is not deflection, it is churn with extra steps. Always keep the path to a person one click away, and treat a customer who escalates after reading three articles as a signal that those articles failed, not as a nuisance. Good deflection feels like help, not a wall. ## A quick deflection checklist - Top ten questions written as findable articles - Help center embedded in the widget, not buried in a footer - Search that returns the right answer for two-word queries - Status page for "is it down?" tickets - Public roadmap/changelog for "when are you adding X?" - Proactive in-product help at known friction points - A grounded AI agent for the routine, with clear escalation - A weekly look at failed searches and articles that are not deflecting ## The payoff for a lean team When deflection works, the math changes. Support volume decouples from headcount, your best people spend their time on the conversations that actually need them, and you get your days back to build the product instead of answering the same question for the hundredth time. It also protects revenue: fast, low-effort help is one of the quietest drivers of retention, which we cover in [how to reduce customer churn](/blog/reduce-shopify-app-churn/). [Convot](/) bundles the whole self-serve layer, a [help center](/help-center/), [status page](/status/), [roadmap and changelog](/portal/), and a grounded [AI agent](/ai/), under one login and one bill, so a small team can deflect like a big one. [Start free](https://app.convot.io/signup) and build the layer that answers for you. --- # AI support quality grading: scoring every conversation URL: https://convot.io/blog/ai-support-quality-grading/ Published: 2026-07-01 Summary: You can't improve support you don't measure. Here's how automatic AI grading scores every conversation, so you find the weak spots without reading every thread. Most support quality reviews are theater. A lead spot-checks a handful of conversations a week, scores them on a rubric, and calls it QA. The other 95% of conversations go ungraded, which means the worst ones, the slow replies, the missed questions, the curt tone, mostly slip through. Automatic grading fixes the coverage problem. ## Why manual QA does not scale Reading conversations is slow, and it is the first thing that gets dropped when the inbox is busy. So QA ends up sampling a tiny, often unrepresentative slice. You learn that your best agent on a calm day does fine. You learn nothing about the 2am thread that pushed a merchant toward churn. ## Grade every conversation, not a sample The alternative is to score every conversation automatically. Convot's [AI quality grading](/ai/) reads each resolved conversation and rates the support, so you get coverage across the whole inbox instead of a hand-picked few. The conversations that score low are the ones worth your attention, surfaced for you instead of hidden in the pile. ## Measure the things that actually matter Good grading is not about politeness scores. It looks at whether the merchant's actual question got answered, whether the response was timely, and whether the issue was resolved or just deflected. Those are the things that move your App Store rating and your churn, so those are the things to measure. ## Turn scores into coaching A grade is only useful if it changes behavior. Use low-scoring conversations as concrete coaching examples, not as a stick. "Here's a thread where the question got missed, here's what good looks like" teaches far better than an abstract rubric. Over time, the patterns in your low scores tell you what to fix systemically, a confusing feature, a missing help article, a gap in your canned responses. ## Tie quality to revenue The highest-leverage move is to weight quality by what is at stake. A mediocre reply to a free-tier tester is a minor issue. The same reply to a [high-MRR merchant](/revenue/) is a revenue risk. When grading and revenue data sit together, you know which quality misses actually cost you money. If you're also using an AI agent to deflect tickets, [What is an AI support agent? (And how to keep it from hallucinating)](/blog/what-is-an-ai-support-agent/) explains the grounding principles that make agent answers worth grading in the first place. Support you measure is support you can improve. Automatic grading just makes "measure everything" finally possible for a small team. See [the AI layer](/ai/), or [start free](https://app.convot.io/signup). --- # The best help desk software for small teams (2026) URL: https://convot.io/blog/best-help-desk-software-small-teams/ Published: 2026-06-29 Summary: Most help desk software roundups are built for enterprises. An honest guide to the best help desk software for small SaaS and software support teams in 2026. Most help desk software roundups are quietly written for enterprise IT teams. They rank tools on ITIL compliance, SLA dashboards, and ticketing configurability, none of which matter much if you're a SaaS team of five people trying to answer customer questions between shipping features. This guide is for small software and SaaS support teams (roughly 2–50 people). The criteria are different: pricing has to work at your size, setup shouldn't require a dedicated support ops person, and the tool ideally covers your inbox, help center, and AI in one place rather than five integrations. ## What small teams actually need from help desk software Before comparing tools, it helps to agree on what matters: - **Affordable, predictable pricing:** per-seat fees that scale with headcount are a tax on growth. Flat pricing or generous tiers matter more than raw feature counts. - **Low configuration overhead:** you don't have a "support ops" person. The tool has to work on day one. - **Channels in one place:** email, live chat, and social in a single inbox, not separate products stitched together. - **Self-serve alongside live support:** a good help center reduces inbound volume. The two should be connected, not bolted together. - **AI that actually works:** an AI agent that makes things up is worse than no AI agent. Look for grounding, source citations, and a clear escalation path. ## Zendesk Zendesk is the category leader for a reason: it handles complexity well. Advanced routing, deep integrations, an enterprise-grade AI layer, and a large partner ecosystem make it the right call for big support operations. **Pros:** Extremely configurable, huge integration library, mature AI, enterprise compliance features. **Cons:** Built and priced for enterprise. Per-agent tiers get expensive quickly, and features you'd expect to be standard often live behind higher plans or paid add-ons. For a team of five, you pay for infrastructure you won't use, and the configuration overhead falls on whoever has the most patience. See our [Zendesk alternatives guide](/blog/best-zendesk-alternatives-small-teams/) if you're actively evaluating replacements. ## Freshdesk Freshdesk competes directly with Zendesk on price and often wins for cost-conscious teams. A generous free plan, solid ticketing across email and chat, and a broad automation library make it a natural choice for teams that want breadth without the enterprise price tag. **Pros:** Strong free tier, wide channel coverage, solid automation, large ecosystem. **Cons:** "Omnichannel" often means separate Freshworks products: Freshchat for live chat sits apart from the core helpdesk product, which can fragment the experience. The UI shows its age in places. AI features have historically been add-on priced, though the product continues to improve. Teams that want a genuinely unified inbox can find the multi-product structure frustrating. ## Help Scout Help Scout's design philosophy is intentional restraint. It looks and feels like email, which makes it easy to adopt: no training, no ticketing jargon, no steep learning curve. Its help center product (Docs) integrates cleanly. **Pros:** Beautifully simple, email-first UX, fast onboarding, well-integrated Docs, genuinely responsive support from the Help Scout team itself. **Cons:** Live chat (Beacon) is functional but not best-in-class. Per-user pricing still scales with headcount, though it's more accessible than Zendesk's. If you need strong live chat, AI resolution, or deep product tooling, you'll hit its limits. A natural choice for teams whose primary channel is email and who value simplicity above all else. ## Intercom Intercom invented the modern in-app messenger and has grown into a full support platform. Fin, its AI agent, is one of the more mature in the market and handles a wide class of support queries well. **Pros:** Best-in-class in-app chat, strong automation, proactive messaging, Fin AI is genuinely capable and battle-tested. **Cons:** Expensive for small teams: per-seat pricing combined with Fin's ~$0.99 per AI resolution (approximate, as of 2026; verify current pricing) means costs compound as usage grows. The feature surface creates real configuration overhead. See our [Intercom alternatives guide](/blog/best-intercom-alternatives-saas-support/) for a broader comparison. Worth it if budget isn't the constraint and you want the premium live-chat experience. ## Front Front reimagines the shared inbox as a collaborative email client, with assignment, comments, SLAs, and analytics layered on top of a familiar email UI. It's popular with customer success and account management teams who manage high-touch relationships. **Pros:** Excellent email collaboration, strong analytics, familiar UX for teams that live in their inbox. **Cons:** More expensive than most alternatives here. Live chat and self-serve help center aren't its strengths. Better suited to high-touch CS than high-volume transactional support. If your primary metric is ticket deflection, Front isn't optimized for that. ## Crisp Crisp punches above its price point for early-stage teams. Live chat, a basic help center, email, team inboxes, and a growing set of automations, at a contact-based price accessible long before you have a real support budget. **Pros:** Affordable, fast to set up, good live chat UX, multilingual support, generous features at the lower tiers. **Cons:** Help center and workflow tooling are simpler than dedicated products. Lacks the depth for teams that need strong escalation flows, quality scoring, or AI resolution. As volume grows, teams often find themselves working around its limits. A good fit for early-stage or bootstrapped teams that want something solid without spending much. ## Convot We built [Convot](/) for exactly this gap: small SaaS and app teams that want a modern support stack without the enterprise complexity or per-seat tax. Shared inbox (live chat, email, and social), help center and knowledge base, Cove AI agent, changelog, product roadmap, quality scoring, and scheduling: one platform, one price. **Pros:** - **No per-seat pricing:** flat platform fee, so adding agents doesn't compound your bill. - **[Cove AI agent](/cove-ai/):** grounded in your docs and account data, cites its sources, and escalates to a human when unsure. At **$0.20 per resolution** (versus Intercom Fin's ~$0.99; check each vendor's current pricing), and free under $1k MRR. - **All-in-one:** inbox, help center, AI, and product tools without the integration overhead. - **Free under $1k revenue:** no credit card required to start. **Cons:** Newer than the incumbents, so the integration ecosystem is still growing. Migration from Crisp is one-click; for Zendesk, Intercom, and Freshdesk we handle the migration manually, free, but it's not a self-serve button. See the full [Convot vs Zendesk comparison](/vs/zendesk/) and [Convot vs Intercom comparison](/vs/intercom/) for deeper breakdowns. ## How to choose | If your team... | Best fit | |---|---| | Needs enterprise-grade routing and compliance | Zendesk | | Wants broad coverage and a strong free tier | Freshdesk | | Is email-first and values simplicity | Help Scout | | Has budget and wants the premium in-app chat experience | Intercom | | Handles high-touch customer relationships via email | Front | | Is early-stage and cost-sensitive | Crisp | | Wants all-in-one, flat pricing, and honest AI | Convot | ## What most roundups miss The feature column tells you what a tool can do. The pricing page tells you what it costs at your current size. The real question is what it costs at 3× or 10× your current size, and whether the tool will still be the right fit. Per-seat pricing is the main variable. A tool that costs $50/month for two agents costs $250/month for ten, before volume or feature upgrades. Flat pricing doesn't have that compounding problem. Check our [pricing page](/pricing/) and [try Convot free](https://app.convot.io/signup): no credit card, free under $1k MRR. --- # The best Intercom alternatives for SaaS support teams (2026) URL: https://convot.io/blog/best-intercom-alternatives-saas-support/ Published: 2026-06-29 Summary: Intercom's per-seat pricing plus Fin's ~$0.99/resolution adds up fast. Here are the best Intercom alternatives for SaaS support teams in 2026. Intercom set the bar for in-app customer support. The live chat experience is polished, the automation is deep, and Fin (their AI agent) is one of the more mature products in the space. But the pricing structure has a sting for small SaaS teams: per-seat licensing stacks with headcount, and Fin's published ~$0.99 per AI resolution means your bill grows as your bot works harder. That combination is what pushes many SaaS support teams to look elsewhere. Here's an honest look at the best Intercom alternatives in 2026. ## What to weigh before switching Intercom is genuinely good at what it does. Before switching, be clear on what you're trading. If you're thinking more broadly about how to build SaaS support that scales, [SaaS Customer Support: A Practical Guide for Lean Teams](/blog/saas-customer-support/) is a useful starting point. - **In-app chat and product tours:** Intercom's messenger is still best-in-class for contextual, in-product conversations. - **Fin's maturity:** the AI agent is well-regarded and has years of training behind it. - **Ecosystem:** deep integrations with CRMs, Salesforce, HubSpot, and data tools. If you rely heavily on proactive messaging or product tours, check whether an alternative covers that before committing. If the core issue is cost or complexity, read on. ## Zendesk Zendesk is the other enterprise-tier standard. Ticketing is its DNA: strong routing, SLAs, and reporting. If you're coming from Intercom because you want more structured ticket management, Zendesk is worth evaluating. **Pros:** Mature ticketing workflows, strong reporting, huge ecosystem, well-understood by support hires. **Cons:** Also per-seat, also add-on-heavy. The AI tier (Zendesk AI) is priced separately, and the total cost can rival Intercom. More of a trade-off in complexity than a step down in price. See our [Zendesk alternatives comparison](/blog/best-zendesk-alternatives-small-teams/) for more. Worth it if you're moving toward a structured ticketing org rather than a live-chat-first model. ## Help Scout Help Scout is the clean, email-first alternative that consistently wins over teams who find Intercom's complexity more burden than benefit. Shared inbox, a Docs help center, and Beacon (their in-app widget), all in a straightforward package. **Pros:** Simple to set up and run, good email workflow, genuine reliability, excellent customer support reputation. **Cons:** Per-user pricing, though friendlier tiers. In-app chat (Beacon) is functional but lighter than Intercom's messenger. Automation and AI are improving but not as deep. Less suited for teams that lean on proactive messaging. A strong choice if your support is mostly reactive and email-heavy. ## Front Front reimagines the shared inbox as a collaborative email client; conversations stay email-shaped, with assignment, comments, and SLAs layered on top. It's popular with customer success teams who handle complex, relationship-driven accounts. **Pros:** Excellent email collaboration, good for account-management-style CS, strong analytics. **Cons:** Expensive. Live chat and self-serve help center are not its core strength. If your team lives in in-product chat, Front won't replace Intercom's messenger well. Worth considering for high-touch CS teams where email is the primary channel. ## Crisp Crisp is the low-cost, all-around-capable alternative that's genuinely underrated. Live chat, email, help center, basic automation, all at contact-based pricing that's accessible for growing teams. **Pros:** Affordable, easy to set up, good live chat UX, no per-seat pricing. **Cons:** AI and automation capabilities are lighter than Intercom or Zendesk. Reporting is simpler. Less suited to teams that need complex escalation flows, quality scoring, or deep CRM integrations. A great fit if you're earlier-stage and Intercom's pricing hit before you needed all of its features. ## Pylon Pylon is newer and specifically built for B2B SaaS teams that support customers via Slack Connect or Microsoft Teams channels. If your enterprise customers expect to reach you in their Slack workspace, Pylon consolidates those channels into a single support interface. **Pros:** Excellent Slack Connect and Teams support, built for B2B relationships, modern interface. **Cons:** Narrower use case, best when your customers actually use Slack Connect with you. Not a full live chat replacement for companies with self-serve or consumer customers. Worth a close look if Slack Connect is central to your enterprise support workflow. ## Convot We built [Convot](/) for SaaS teams that want a modern all-in-one support stack without the enterprise pricing structure. Shared inbox (live chat, email, social), full help center and KB, [Cove AI agent](/cove-ai/), product roadmap, changelog, scheduling, and quality scoring, all in one platform. **Pros:** - **No per-seat pricing:** flat platform fee. Your bill doesn't change when you hire a second support rep. - **[Cove AI](/cove-ai/)** resolves queries grounded in your docs and account data, cites sources, and escalates when it's unsure. At **$0.20 per resolution** (about a fifth of Fin's published ~$0.99), it's meaningfully cheaper at scale. - **Free under $1k MRR:** no credit card required. Full platform, not a feature-limited trial. - **All-in-one:** inbox, help center, AI, changelog, roadmap, and scheduling without a separate tools budget. **Cons:** Newer platform, so the ecosystem of third-party integrations is still growing. Migration from Intercom is handled manually by us, free; it's not a one-click button. One-click import is available for Crisp. See the full [Convot vs Intercom comparison](/vs/intercom/) for a detailed breakdown. ## How to choose | If you... | Consider | |---|---| | Need structured ticketing and strong reporting | Zendesk | | Want simple, reliable email-first support | Help Scout | | Handle complex, relationship-driven accounts | Front | | Are cost-sensitive and want easy live chat | Crisp | | Support enterprise customers via Slack Connect | Pylon | | Want all-in-one, flat pricing, fair AI rates | Convot | ## On AI resolution pricing This is worth calling out directly. When your AI agent resolves 1,000 conversations a month, the difference between $0.20 and $0.99 per resolution is $790/month, almost $10k a year. That's not a rounding error. Before committing to any tool with per-resolution AI pricing, run that math for your actual volume. For a deeper look at how AI agents work and what makes them reliable, see [What is an AI support agent? (And how to keep it from hallucinating)](/blog/what-is-an-ai-support-agent/). Cove's $0.20/resolution pricing, and the fact that it's free under $1k MRR, is why we built it that way. [Check our pricing](/pricing/) and [try it free](https://app.convot.io/signup). --- # Crisp vs Intercom for Shopify app developers: which to pick URL: https://convot.io/blog/crisp-vs-intercom-for-shopify-apps/ Published: 2026-06-29 Summary: Crisp and Intercom are both solid chat tools. But if you build Shopify apps, the right choice depends on price, depth, and one thing neither of them does. If you are choosing a support tool for your Shopify app, Crisp and Intercom are two of the names you will weigh. Both are mature and capable. Here is how they actually differ for an app team, and the question that matters more than either. ## Crisp: affordable and lean Crisp is the budget-friendly option. You get live chat, a shared inbox, a help center, and a chatbot builder at a price that works for a small team. It is generous on its lower tiers and does not nickel-and-dime you. The tradeoff is that it is a general-purpose tool: it knows nothing about Shopify, and its automation and reporting are lighter than Intercom's. ## Intercom: powerful and pricey Intercom is the heavyweight. Its messaging, automation, AI agent, and reporting are deep, and it scales to large support orgs. The tradeoff is cost and complexity: pricing climbs quickly with seats and resolutions, and the feature surface is far more than a three-person app team needs. You pay for a platform built for companies much bigger than yours. ## The question neither answers Here is the catch for a Shopify app team: neither Crisp nor Intercom can tell you that the merchant messaging you pays $200 a month, is on your Pro plan, and is worth $1,800 in lifetime value. They were built for generic businesses, not for the Shopify Partner ecosystem. So you end up running a separate revenue dashboard and mentally stitching it to your inbox. ## Where Convot fits Convot covers the same support essentials, live chat, shared inbox, help center, mobile app, and then connects to your Shopify Partner account and puts each merchant's [MRR, plan, and LTV beside the conversation](/revenue/). It also bundles a changelog, roadmap, status page, and scheduling, so you replace several tools with one. And the SDK is Crisp-compatible, so [migrating from Crisp](/migrate/) is close to copy-paste. If you want the full breakdowns: [Convot vs Crisp](/vs/crisp/) and [Convot vs Intercom](/vs/intercom/). ## How to choose Pick Crisp if budget is the only constraint and you do not need Shopify context. Pick Intercom if you are a large org that needs its depth and can afford it. Pick Convot if you build Shopify apps and want support that knows what each merchant is worth, without paying for five separate tools. [Start free](https://app.convot.io/signup) with your first app. --- # The best Zendesk alternatives for small support teams (2026) URL: https://convot.io/blog/best-zendesk-alternatives-small-teams/ Published: 2026-06-29 Summary: Zendesk is powerful but built for large orgs. Here's an honest comparison of the best Zendesk alternatives for small SaaS support teams in 2026. Zendesk is excellent software, if you have a dedicated support ops team to configure it, a budget to match its per-agent tiers, and the patience to evaluate add-ons that unlock features you assumed were included. For a small SaaS support team of one to five people, that calculus rarely works out. If you're evaluating Zendesk alternatives, here's an honest look at the tools worth considering in 2026. ## What small teams actually need The criteria change when you're small. If you're also comparing across more tools, [The best help desk software for small teams (2026)](/blog/best-help-desk-software-small-teams/) covers the wider landscape. - **Simplicity over configurability:** you don't have a Zendesk admin. You need something that works on day one. - **Flat or predictable pricing:** per-agent fees that scale with headcount hurt as you hire. - **All-in-one coverage:** a separate ticketing tool, help center, and AI agent adds integration overhead you don't want. - **Speed:** small teams answer fast or they lose customers. Mobile apps and live chat matter more than SLA dashboards. ## Help Scout Help Scout is the most natural first stop for teams leaving Zendesk. It's genuinely simple, with a shared inbox that feels like email rather than a ticketing system, solid help-center tooling (Docs), and a reputation for good customer support itself. **Pros:** Clean UI, low learning curve, good email-first workflow, reasonable pricing. **Cons:** Live chat (Beacon) is functional but limited compared to dedicated chat tools. The AI features are improving but newer. Per-user pricing still scales with headcount, though it's friendlier than Zendesk's tiers. A good fit if your support is primarily email-based and you want a polished, focused tool. ## Freshdesk Freshdesk targets the same Zendesk-is-too-expensive audience directly, with a free plan and competitive paid tiers. It covers ticketing, live chat, phone, and automation, and the feature breadth is impressive. **Pros:** Generous free tier, wide channel coverage, strong automation, large ecosystem. **Cons:** The UI can feel busy and dated in places. "Omnichannel" features often live behind separate Freshdesk products (Freshchat for live chat, Freshservice for IT), which fragments the stack. AI features are improving but have historically been add-on priced. A solid choice if you need breadth and the free plan covers your volume. ## Front Front is built around shared inboxes in email clients; it keeps conversations in an email-like flow and layers assignment, collaboration, and SLAs on top. Popular with customer success and account management teams. **Pros:** Excellent email collaboration, strong for teams that live in their inbox, good analytics. **Cons:** More expensive than the alternatives here. Live chat and self-serve help center are not its strengths. Better suited to high-touch CS than high-volume support. Worth considering if your team's primary channel is email and you want deep inbox collaboration. ## Intercom Intercom invented the modern in-app chat category and has grown into a full support platform with Fin, its AI agent. It's genuinely powerful, with strong product tours, proactive messaging, and a capable bot. **Pros:** Best-in-class in-app chat experience, strong automation, Fin AI agent is mature and well-regarded. **Cons:** Expensive for small teams: per-seat pricing plus Fin's ~$0.99 per AI resolution means costs rise as the product works. Feature depth creates configuration overhead. See our full [Zendesk vs Intercom comparison](/vs/intercom/) for how the two stack up. Worth it if you have budget and want the premium experience. Heavy if you're small and cost-sensitive. ## Crisp Crisp punches above its price point. Live chat, a basic help center, email, and a growing set of automations, all at a contact-based price that's accessible for early-stage teams. **Pros:** Affordable, easy to set up, good live chat UX, multilingual. **Cons:** Help center and ticketing tooling are functional but simpler than dedicated tools. Lacks the deeper workflow automation or AI resolution capabilities of the higher-tier tools. Less suited for teams that need strong escalation flows or quality scoring. A good fit if you're early-stage and want something simple that won't break the budget. ## Convot We built [Convot](/) for small SaaS and app teams that want a modern support stack without the enterprise tax. Instead of separate tools for inbox, help center, and AI, it's one platform: shared inbox (live chat, email, and social), a full help center and KB, Cove AI agent, changelog, product roadmap, scheduling, and quality scoring. **Pros:** - **No per-seat pricing:** flat platform fee. Adding a second agent doesn't change your bill. - **[Cove AI agent](/cove-ai/)** resolves support queries grounded in your docs and account data, cites sources, and escalates when unsure. At **$0.20 per resolution** it's a fifth of Intercom Fin's ~$0.99, and free under $1k MRR. - **All-in-one:** inbox, help center, AI, and product tools in one place, not bolted together. - **Free under $1k revenue:** no credit card required to start. **Cons:** Newer than the incumbents, so ecosystem integrations are still growing. One-click migration from Crisp; for Zendesk migrations we handle it manually, free, but it's not a self-serve button. See the full [Convot vs Zendesk comparison](/vs/zendesk/) for a deeper breakdown. ## How to choose | If you... | Consider | |---|---| | Want the simplest email-first inbox | Help Scout | | Need breadth and a free plan | Freshdesk | | Live in email and need deep collaboration | Front | | Have budget and want premium in-app chat | Intercom | | Are early-stage and cost-sensitive | Crisp | | Want all-in-one, flat pricing, and AI at a fair rate | Convot | ## The real cost of Zendesk The sticker price is only part of the picture. Per-seat licensing means every new hire adds to the bill. Features like advanced AI, sandboxing, or reporting are often gated behind higher tiers or add-ons. For a small team, that total cost of ownership is what drives the switch. [Zendesk pricing explained (2026): what you'll actually pay](/blog/zendesk-pricing-explained-2026/) breaks down exactly how the tiers stack up. Check our [pricing page](/pricing/) to see how Convot compares, and [try it free](https://app.convot.io/signup): no credit card, free under $1k MRR. --- # What is an AI support agent? (And how to keep it from hallucinating) URL: https://convot.io/blog/what-is-an-ai-support-agent/ Published: 2026-06-29 Summary: An AI support agent can resolve customer questions automatically, but hallucination is the top reason teams distrust AI customer support. Here's how grounding and abstain logic make the difference. An AI support agent is software that answers customer support questions automatically: not by routing tickets or suggesting macros, but by reading your documentation and generating real answers. Used well, it resolves the questions your team answers fifty times a day, frees up human time for the harder ones, and stays available at 3 a.m. without a staffing cost. Used carelessly, it confidently tells customers the wrong thing and erodes trust faster than no AI at all. The gap between those two outcomes comes down to one thing: grounding. ## What an AI support agent actually is At its core, an AI support agent combines a large language model with a retrieval layer over your own content: your help center, knowledge base, product documentation, and increasingly, live account data like orders, bookings, or subscription status. When a customer asks a question, the agent: 1. Retrieves relevant content from your knowledge base 2. Generates a response based on that content, not general internet knowledge 3. (Ideally) cites the sources it used, so the customer and your team can verify the answer 4. Escalates to a human agent when the question is outside the scope of its knowledge This is fundamentally different from a generic chatbot. A generic chatbot is trained on internet data and will try to answer anything, often by sounding plausible while being factually wrong. A properly grounded AI support agent works only from what you've told it, and should know when to say "I don't know." The distinction matters because customers assume you stand behind every answer your support channel gives, regardless of whether a human or an AI wrote it. ## The #1 risk: hallucination Hallucination is the term for when an AI model generates a confident answer that isn't supported by the source material, or that's simply invented. In AI customer support, it looks like: - Quoting a refund policy that doesn't exist - Citing a feature your product doesn't have - Giving the wrong steps for a settings change - Confidently answering a billing question with made-up numbers Teams that deploy AI support agents and then turn them off usually cite this as the reason. One wrong answer, especially a visible one, creates more support work than it saved and damages the customer relationship in the process. The problem isn't the technology itself; it's how the agent is configured and constrained. ## How grounding prevents hallucination Grounding means the agent's answers are anchored to specific documents in your knowledge base, and any claim in the response maps back to a retrievable source. A well-grounded agent: - **Only uses your content:** not the open internet, not training data from other companies' docs. If it isn't in your knowledge base, the agent doesn't know it. - **Cites its sources:** each answer links back to the specific help article or document it drew from. Customers can verify; your team can audit when something is wrong. - **Stays in scope:** if a question doesn't match anything in the knowledge base above a confidence threshold, it doesn't guess. It flags the gap. A second layer of protection is **account data grounding**. Many support questions aren't documentation questions at all: "where is my order?", "what plan am I on?", "when is my next booking?" An agent that can only read your docs will escalate all of these. An agent that can also securely read live account data (via an API or an MCP connection) can answer them accurately, because it's reading the actual source of truth rather than inferring from general knowledge. ## Abstain and escalate: the safety net that matters most Even a well-grounded agent will encounter questions it can't reliably answer. The correct behavior in those cases is to **abstain**: to explicitly not answer, and hand off to a human instead. An agent that always generates a response, even when uncertain, will hallucinate. An agent with a calibrated confidence threshold will recognize when it's outside its reliable range and say so. This is the difference between: - *"Your order shipped on June 3rd."* Confidently wrong, with no way for the customer to know that. - *"I couldn't find a reliable answer to this in our documentation. I'm passing this to our team."* An honest escalation that preserves trust. **Suggest mode** is an underrated starting point. Instead of sending AI-generated answers directly to customers, the AI drafts a response for a human agent to review and send. This builds confidence in the system, catches mistakes before they reach customers, and lets you verify accuracy at scale before switching to fully automated responses. Many teams start here and graduate to autonomous mode over time. ## Understanding AI support pricing Two pricing models dominate the market: **Per-seat pricing** bundles the AI into a platform fee paid regardless of how much the AI actually resolves. If the AI handles very little, you're paying for unused capacity. If it handles a lot, the unit economics improve. **Per-resolution pricing** charges for each query the AI successfully resolves. Intercom's Fin charges approximately $0.99 per resolution (as of 2026; verify current pricing on their site). The model aligns costs with value, but the numbers add up: 2,000 AI resolutions a month is roughly $2,000 with that pricing. This model is sometimes called **outcome-based pricing**, and it's increasingly common in AI customer support. Worth modeling your expected resolution volume before committing: the bill can look very different at different scales. Some platforms also offer a **free tier** while you're early-stage, which means you can validate whether AI support actually works for your specific support content before spending anything. ## How Convot's Cove approaches this [Cove](/cove-ai/) is Convot's AI support agent. It was built around the grounding-first principle from day one: - **Grounded in your docs and account data:** Cove draws from your help center and can connect to live account data via a secure API or MCP server. It doesn't guess from general knowledge. - **Source citations on every answer:** each response links back to the specific article or data source it used. Customers see exactly where the answer came from. - **Abstain and escalate logic built in:** when Cove isn't confident, it flags the conversation for a human agent rather than generating an uncertain answer. - **Suggest mode available:** start with a human reviewing every AI draft before it's sent, then graduate to autonomous when you're ready. - **$0.20 per resolution:** compared to Fin's ~$0.99, that's roughly a fifth of the cost at similar volume. And [Convot is free under $1k MRR](/pricing/), no credit card required. If you're evaluating AI support agents and concerned about hallucination risk, the most important question to ask any vendor is: *"What does your agent do when it doesn't know the answer?"* The answer tells you more than any feature list. See also: [best Intercom alternatives](/blog/best-intercom-alternatives-saas-support/) if you're comparing Fin to other options, and [try Convot free](https://app.convot.io/signup) to see Cove in action on your own help center. --- # Zendesk pricing explained (2026): what you'll actually pay URL: https://convot.io/blog/zendesk-pricing-explained-2026/ Published: 2026-06-29 Summary: Zendesk pricing is per-seat, tiered, and add-on-heavy. Here's how the cost structure works, and what to compare it against when evaluating alternatives. Zendesk is transparent about its pricing page. What's less obvious is how the structure works, and how the total bill behaves as your team and your AI usage grow. This post explains the mechanics, not made-up numbers. (All figures are approximate as of mid-2026. Zendesk's exact pricing changes; check the vendor for current rates before making any decisions.) ## How Zendesk's plan structure works Zendesk's core product is the **Suite**, a bundle that combines the ticketing system, help center, live chat, and voice into one SKU. The Suite comes in tiers (Team, Growth, Professional, and Enterprise, with naming subject to change). Each tier is priced **per agent, per month**. That "per agent" model is the most important thing to understand. Every support hire is a line item. A team of five at the mid-tier costs roughly five times a team of one at that same tier. If you're planning to grow headcount, model that trajectory before committing. **Annual vs. monthly billing:** Like most SaaS tools, Zendesk discounts for annual commitments. The gap between monthly and annual billing can be meaningful, worth calculating for your budget cycle. ## What each tier actually gates Lower tiers cover the fundamentals: ticketing, email, basic chat, and a help center. Higher tiers progressively unlock: - **Reporting and analytics:** deeper dashboards and custom reports are typically Professional/Enterprise features. - **SLA management:** more advanced SLA policies unlock at higher tiers. - **Custom roles and permissions:** granular access control is an upper-tier feature. - **Sandbox environments:** usually Enterprise only. The pattern is consistent: features you'd expect to be standard in enterprise software are gated behind tier upgrades. ## Add-ons: where the real costs hide On top of the base tier, Zendesk offers add-ons, including AI features. This is the part that catches teams off-guard. **AI and automation:** Zendesk's AI capabilities (automated triage, AI-assisted replies, bot conversations) have historically been priced separately from the base suite, either as an add-on bundle or as usage-based fees on top of the tier price. This means your per-seat quote may not include the AI features that made Zendesk attractive in the first place. **Per-resolution AI pricing:** This is worth comparing across the market. Intercom, for example, publishes their Fin AI agent at approximately **$0.99 per resolution**, one of the few hard public numbers available. Zendesk's AI pricing structure varies and can include both per-seat components and usage-based components. Ask specifically about the per-resolution or per-usage cost when you get a quote. ## The real cost model: seat × tier × time Here's the mental model for estimating your true Zendesk cost: ``` Monthly cost = (agents × per-agent tier price) + AI/automation add-on + any usage fees ``` At three agents on a mid-tier plan, plus an AI add-on, you're likely looking at a materially different number than the headline per-seat figure suggests. At ten agents plus full AI, the number can be substantial. The key dynamics: 1. **Headcount scales the bill linearly.** Every new hire is a multiplier on your base rate. 2. **AI is often gated or add-on priced.** The features that reduce ticket volume cost extra. 3. **Annual lock-in.** Committing annually saves money but reduces flexibility if your needs change. None of this is hidden; it's just not front-and-center on the pricing page. ## Intercom pricing follows a similar pattern It's worth mentioning Intercom here because the two tools are frequently compared ([we have a dedicated comparison](/vs/intercom/)). Intercom's pricing structure is also per-seat and tiered, with one particularly visible data point: **Fin, their AI agent, is published at approximately $0.99 per resolution.** That's a meaningful figure when you run the math. At 500 AI resolutions per month, that's ~$495 in AI costs alone, on top of per-seat licensing. At 2,000 resolutions: ~$1,980. For teams with high AI usage, the per-resolution model turns the AI (the thing reducing your support burden) into a significant cost driver. ## A different model: flat pricing + $0.20 per resolution [Convot](/) was built around a different pricing philosophy, specifically because per-seat and per-resolution models penalize growth. The structure: - **Flat platform fee, no per-seat charges.** A five-agent team pays the same as a one-agent team. Hiring doesn't change the bill. - **[Cove AI](/cove-ai/) at $0.20 per resolution:** about a fifth of Intercom Fin's published ~$0.99. Cove resolves queries grounded in your actual docs and account data, cites sources, and escalates when it's unsure rather than guessing. - **Free under $1k MRR.** The full platform (shared inbox, help center, AI agent, changelog, roadmap, scheduling, quality scoring) at no cost until you're making money. At 1,000 resolutions/month: Cove = $200 vs. Fin's ~$990. The delta widens as volume grows. See the [Convot vs Zendesk comparison](/vs/zendesk/) for a side-by-side of features and structure. For a broader look at the alternatives, read our [Zendesk alternatives post](/blog/best-zendesk-alternatives-small-teams/). ## How to evaluate any support tool's pricing Before signing: 1. **Model your team size in 12 months**, not today. Multiply per-agent fees by that number. 2. **Ask specifically about AI pricing:** is it included, add-on, or usage-based? 3. **Request an annual vs. monthly comparison**, and check whether you need the flexibility of monthly. 4. **List the features you need on day one** and verify which tier gates them. For most small SaaS teams, the tools with flat or transparent pricing structures simply carry less risk than the per-seat-plus-add-on models, even if the headline per-seat price looks competitive. [Check Convot's pricing](/pricing/) and [try it free](https://app.convot.io/signup): no credit card, free under $1k MRR. --- # Book onboarding calls without leaving your support chat URL: https://convot.io/blog/book-onboarding-calls-from-support-chat/ Published: 2026-06-26 Summary: A pasted scheduling link kills momentum. Here's how to turn a support conversation into a booked call inside the widget, and when a call beats a ticket chain. Sometimes a support thread is going nowhere and a five-minute call would solve everything. But the moment you paste "here's my Calendly," you break the momentum: the merchant leaves the chat, lands on a third-party page with someone else's branding, and half the time never books. Here is how to do it better. ## A call beats a ticket chain when the problem is fuzzy Text is great for clear, contained questions. It is terrible for "it's not working and I can't explain why." When a problem is fuzzy, high-stakes, or emotional, a short call resolves it faster than ten back-and-forth messages, and it converts better, because the merchant feels heard. ## Keep the booking inside the conversation The fix is to let the merchant book without leaving the chat. Convot's [in-chat scheduling](/scheduling/) sends a booking card right in the conversation: the merchant picks a time, gets a Google Meet link and a calendar invite, and never sees a "Powered by" badge for someone else. No context switch, no pasted link, no momentum lost. It is a [Calendly alternative](/vs/calendly/) that lives where your support already happens. ## Use it for the moments that matter You do not want to offer a call on every ticket. Use it deliberately: onboarding a high-value merchant, saving a frustrated one, walking through a complex setup, closing an upgrade. When you can see the merchant's [revenue](/revenue/) beside the chat, you know which conversations are worth a calendar slot and which are better handled with a help article. ## Reduce the no-shows A booked call only helps if the merchant shows up. Automatic reminders and a calendar invite with the meeting link cut no-shows. Convot handles the .ics file, the Meet link, and branded reminder emails, so the call actually happens. ## Turn the call into retention The best onboarding and save calls do double duty: you solve the immediate problem and you learn what nearly made the merchant leave. That feedback feeds straight back into your product and your help center, so the next merchant does not hit the same wall. A scheduled call at the right moment is one of the highest-converting support moves there is. The trick is making it frictionless. See [scheduling](/scheduling/), or [start free](https://app.convot.io/signup). --- # Catch frustrated merchants before they leave a 1-star review URL: https://convot.io/blog/catch-frustrated-merchants-before-bad-reviews/ Published: 2026-06-24 Summary: A one-star review usually starts as a frustrated support thread you didn't notice in time. Here's how to spot rising frustration and step in while it's still fixable. Almost no one leaves a one-star review out of nowhere. It starts as a support conversation that went sideways, a question that took too long, a tone that got sharper with each reply, a problem that did not get solved. By the time the review appears, the moment to fix it has passed. The whole game is catching that frustration while the thread is still open. ## Frustration is a signal you can read In a busy inbox, the conversation turning hostile looks the same as every other open ticket until it explodes. But the signs are there: shorter replies, stronger language, repeated follow-ups, "this is the third time I've asked." A human catches it when they happen to reread the thread. The problem is catching it across every conversation, every day. ## Let the system watch for you Convot reads every conversation as it happens and escalates the ones turning hostile in real time. Its [frustration escalation](/ai/) flags a heated thread and can alert a manager, so someone steps in while the merchant is still talking to you instead of typing a review. You are not relying on an agent to notice; the system surfaces the conversation that needs a human now. ## Prioritize by what's at stake A frustrated merchant paying you $200 a month is a five-alarm fire. A frustrated free-tier tester is a normal ticket. When you can see [revenue beside the conversation](/revenue/), you know which frustrated thread to drop everything for. Frustration plus high MRR is the exact combination worth interrupting your day for. ## Make the recovery count Catching frustration only matters if the recovery is good. Acknowledge the problem directly, skip the canned apology, fix the actual issue, and where it makes sense, offer something, a credit, a fix timeline, a quick call. A genuinely good recovery does not just prevent a one-star review; it sometimes earns a five-star one, because you showed up when it mattered. ## The compounding effect on your rating Your App Store rating is the sum of a thousand small support moments. Every frustrated thread you catch and turn around is a one-star review that never happened and an install that keeps trusting you. Catching frustration is the defensive half of rating management; the offensive half is [earning more 5-star reviews](/blog/how-to-get-more-shopify-app-reviews/) by asking at the moment a merchant just felt the value. Over months, that is the difference between a rating that climbs and one that bleeds. See how [the AI layer](/ai/) works, or [start free](https://app.convot.io/signup). --- # Mantle Is Winding Down: How to Move Your Revenue Dashboard Before Aug 14 URL: https://convot.io/blog/mantle-winding-down-move-revenue-dashboard/ Published: 2026-06-24 Summary: Mantle's non-billing services end August 14, 2026. A calm, practical migration guide for Shopify app teams: what actually needs to move, what rebuilds itself from the Shopify Partner API, and how to sequence it before the deadline. On June 16, Mantle announced it is winding down. If you build Shopify apps, Mantle was probably the tool you opened every morning to see MRR, plans, LTV, and churn in one place. Losing it stings, and the timeline is short. So this is not a pitch. It is the calm version of "what do I actually do now," written by a team in the same ecosystem. First, the dates, because they set everything else: - **August 14, 2026** is the real deadline. Non-billing services (analytics, help desk, CRM, email, APIs, MCP) go dark 60 days after the announcement. - **September 30, 2026** is the billing runway. Mantle steers billing accounts to Shopify's native App Pricing. The instinct is to panic about exporting everything. You mostly do not need to, and here is the part that takes the pressure off. ## Your revenue dashboard was never really Mantle's data The MRR, plan, LTV, and churn view you check daily was built from the Shopify Partner API. Mantle read that API and drew you a dashboard. The numbers live in Shopify, not in Mantle. That means the one thing you are most afraid of losing is portable by definition: any tool wired to the Partner API can regenerate the same view. There is nothing to export, no CSV to wrangle, no history trapped behind a sunset date. You reconnect the Partner API somewhere new and the dashboard rebuilds itself. That reframes the whole migration. The scary part is not the part that is actually at risk. ## The honest migration checklist, hardest first Not everything Mantle did needs a one-to-one replacement. Here is what to handle, in priority order: **1. Help-desk history (do this first).** This is the one piece with no clean re-sync, because it is conversation data Mantle stored, not Shopify data. If you ran support through Mantle, request an export from their support team now, while they are still staffed for it. Leaving this until early August is the actual risk in the whole migration. **2. Contacts and CRM.** Export your customer and contact records to CSV while the account is live. Most support and CRM tools import a CSV directly. **3. The revenue dashboard.** As above: nothing to export. Reconnect the Shopify Partner API wherever you move and it rebuilds. Prioritize this lower precisely because it is not at risk. **4. Billing.** Do not go hunting for a Mantle billing replacement. Mantle's own recommendation is Shopify's native App Pricing, and for most teams that is the right answer. Shopify owns this now. **5. Email, affiliates, App Store analytics.** These are separate jobs that happened to live under one roof. If you used them, they move to dedicated tools, not to a single all-in-one. Anyone telling you one product replaces all of Mantle is selling you a worse version of each piece. ## How to sequence it before August 14 You have about eight weeks. A calm order: - **This week:** request your help-desk history export from Mantle, and export contacts to CSV. These are the time-sensitive, staffing-dependent steps. - **Next two weeks:** pick where the revenue view and support live going forward, reconnect the Partner API, and confirm the dashboard rebuilds the way you expect. - **By early August:** move billing to Shopify's native App Pricing (you have until Sep 30, but do not leave it to the last week), and point your changelog, status page, and roadmap to their new homes. - **Buffer the final week.** Do not plan to migrate anything in the last days before the 14th. ## Where this lands for support teams If you are rebuilding the revenue dashboard anyway, it is worth asking where it should live. The reason we built [Convot](/vs/mantle/) is that a revenue number is most useful at the moment a merchant is at risk, not in a dashboard you check once a day. Convot pulls the same Shopify Partner API Mantle did, and puts each merchant's MRR, plan, LTV, and [churn risk](/blog/reduce-shopify-app-churn/) right beside the support conversation. When a merchant worth $300 a month goes quiet or sounds frustrated, you see it while you can still do something. After an uninstall, its [AI reads the history](/blog/when-a-merchant-uninstalls-your-shopify-app/) and tells you whether support was the cause. To be upfront about what it is not: Convot is not a billing engine, an email platform, an affiliate tool, or an App Store ranking product. Billing goes to Shopify. We would rather say that than pretend to replace everything. ## The short version The data you are most afraid of losing, the revenue dashboard, is the data that is safest, because it lives in the Shopify Partner API and rebuilds anywhere. The thing actually at risk is your help-desk history, so export that this week. Move billing to Shopify native. Handle the rest calmly, with a buffer before August 14. Mantle set a high bar for app-partner tooling, and the move off it is less of a fire drill than it feels like today. If you want the revenue view and your support inbox in one place, here is [how the move to Convot works](/vs/mantle/). And if you just want a second pair of eyes on your migration plan, no pitch, we are happy to help. --- # Shopify app onboarding: turning installs into paying merchants URL: https://convot.io/blog/shopify-app-onboarding-installs-to-paying/ Published: 2026-06-22 Summary: Most churn happens in the first two weeks, before a merchant ever pays. Here's how to design onboarding and support that turns installs into active, paying merchants. The biggest leak in most Shopify apps is not late-stage churn. It is the install that never activates. A merchant installs, hits a confusing step, drifts away, and uninstalls a week later without ever experiencing the value of your app. Fixing that leak is the highest-return work you can do. ## The first session is everything A merchant decides whether your app is worth their time in the first session. If they reach the "aha" moment, the point where the app visibly does the thing they installed it for, they stick. If they get stuck before that, they leave. So map your onboarding backward from that moment and remove every step between the install and it. ## Be present at the friction points Onboarding flows break in predictable places: the permission screen, the first configuration, the theme integration. Put live chat on those exact screens so a stuck merchant can ask instead of giving up. Convot lets you embed a [branded chat widget](/live-chat/) right where the friction is, so the question gets asked instead of becoming a silent uninstall. ## Answer the first message fast The first support message from a new merchant is the most important one they will ever send. A fast, helpful reply in the first session converts a wobbling trial into an active user. A slow one confirms their suspicion that the app is not worth it. This is where a low first-response time pays for itself directly in activation. ## Reach out proactively Do not wait for the merchant to ask. If you can see that a merchant installed but never completed setup, reach out. A single proactive message, "noticed you started setting up, anything I can help with?", recovers a meaningful share of stalled installs. ## Know who is worth the extra effort Some installs are worth a personal onboarding call; most are not. When you can see a merchant's [plan and revenue potential](/revenue/) beside the conversation, you know where to spend your limited time. A high-value merchant stuck in onboarding is worth a [scheduled call](/scheduling/); a free tester is worth a help article link. Win the first two weeks and you fix the leak that quietly caps every app's growth. [Start free](https://app.convot.io/signup) and put support where onboarding breaks. --- # Why your Shopify app needs a status page URL: https://convot.io/blog/status-page-for-your-shopify-app/ Published: 2026-06-19 Summary: When your app goes down, merchants flood your inbox. A status page turns a support fire into a calm, public update. Here's how to run one for a Shopify app. The worst time to handle support is during an outage. Your app is down, every affected merchant messages you at once, and you are trying to fix the problem and answer fifty identical "is it just me?" messages at the same time. A status page solves exactly this. ## Turn a fire into a single update A status page lets you say "we know, here is what is happening, here is the ETA" once, publicly, instead of fifty times in your inbox. Merchants who would have opened a ticket check the page instead. Your inbox stays calm enough that you can actually fix the problem. ## Automate the detection A status page you have to remember to update during a crisis is a status page that stays green while everything is on fire. The useful version watches your endpoints and creates an incident automatically when something goes down. Convot ships a [status page with built-in uptime monitoring](/status/) that auto-creates incidents when a URL stops responding, so the page is honest even when you are heads-down debugging. ## Put it where merchants already are A status page on a domain nobody bookmarks does not help. The status needs to be visible where merchants already look for help, on your own domain and inside your support widget. That way a merchant about to message you sees the active incident first. It is a built-in [Statuspage alternative](/vs/statuspage/) without the separate subscription. ## Build trust over time The quiet benefit of a status page is trust. A public 90-day uptime history says "we are reliable and we are honest when we are not." For a Shopify app, where merchants are betting their store's workflow on you, that credibility matters. Merchants forgive an outage that is communicated. They churn over an outage that is hidden. ## Keep the post-mortem short and human When an incident resolves, write a short, plain-language note: what happened, what you did, what you are changing. No corporate hedging. That note turns an outage from a reason to leave into a reason to trust you more. A status page is cheap insurance. The one hour it takes to set up pays for itself the first time your app has a bad morning. See [the status page](/status/), or [start free](https://app.convot.io/signup). --- # Running a public roadmap your merchants vote on URL: https://convot.io/blog/public-roadmap-for-your-shopify-app/ Published: 2026-06-17 Summary: A public roadmap turns feature requests into retention. Here's how to run one for your Shopify app without it becoming a graveyard of ignored ideas. When a merchant requests a feature, they are telling you they care enough about your app to want it to be better. A public roadmap turns that signal into a retention loop: they suggest, they vote, you ship, they stay. Done badly, it becomes a graveyard of ignored ideas. Here is how to do it well. ## Let merchants vote, not just submit A list of requests is noise. A list of requests with votes is a prioritized backlog built by the people paying you. Voting tells you what to build next and gives merchants a sense of ownership. The ones who vote and see their idea ship are the ones who renew. ## Close the loop when you ship The single biggest mistake teams make is shipping a requested feature and never telling the people who asked. Email the voters when their idea goes live. That one message does more for loyalty than a month of marketing, because it proves you listened. Convot's [roadmap and changelog](/portal/) do this with one click: merchants vote on the roadmap, and when the request ships, the voters get an email. It is a [Canny alternative](/vs/canny/) that lives inside the same widget as your support, so merchants vote where they already chat with you. ## Tie votes to revenue Not all votes are equal. A feature requested by three merchants paying you $200 a month is a different priority from one requested by thirty free-tier testers. When your roadmap is connected to your [revenue data](/revenue/), you can see the MRR behind each request and prioritize the revenue, not just the vote count. ## Keep it honest A roadmap is a promise, so do not over-promise. Use a small number of public columns, under consideration, planned, shipped, and move things deliberately. Merchants forgive a slow roadmap. They do not forgive a roadmap full of "planned" items that never move. ## Pair it with a changelog The roadmap shows what is coming; the changelog shows what shipped. Together they tell a merchant your app is alive and improving, which is one of the quietest but strongest retention signals there is. A public roadmap is not a vanity feature. It is a structured way to let your best merchants tell you how to keep them. See [the product portal](/portal/), or [start free](https://app.convot.io/signup). --- # Building a help center for your Shopify app URL: https://convot.io/blog/build-a-help-center-for-your-shopify-app/ Published: 2026-06-15 Summary: A help center is the cheapest support you'll ever run. Here's how to structure one for a Shopify app, what to write first, and how to keep it deflecting tickets. Every merchant who finds the answer themselves is a ticket you never have to handle. A good help center is the highest-leverage support investment a small app team can make: write an article once, deflect the same question hundreds of times. ## Write the questions you already answer Do not start with a content plan. Start with your inbox. The questions you answer most often are exactly the articles you should write first: setup steps, the one feature everyone misconfigures, billing questions, the error people hit most. Ten articles that cover your top ten questions will deflect more tickets than fifty written in a vacuum. ## Structure it for scanning, not reading Merchants do not read help articles. They scan them while half-frustrated. Use short sections, clear headings, numbered steps, and screenshots. Put the answer near the top. Link related articles so one good page leads to the next. ## Put it where merchants get stuck A help center on a separate domain that nobody visits does not deflect anything. The articles need to be searchable at the moment of friction, inside the support widget, on your app screens, and on your own domain for SEO. Convot puts your [knowledge base on your domain and inside the chat widget](/help-center/), searchable in the conversation before a merchant ever opens a ticket. It is a built-in [Zendesk Guide alternative](/vs/zendesk/), without the separate subscription. ## Support more than one language If your app sells globally, your help center should too. Multi-language articles let a merchant read the answer in their own language, which deflects tickets you would otherwise handle slowly through translation. ![A help center article with a "was this helpful?" prompt](/shots/help-article.png) ## Measure deflection, then improve A help center is not "done." Watch which articles get viewed, which searches return nothing, and which questions still reach your inbox despite an article existing (usually the article is hard to find or unclear). Add the missing articles, rewrite the unclear ones, and your deflection rate climbs over time. ## The compounding effect The best part is that a help center compounds. Every article you write keeps working while you sleep, deflecting tickets and ranking in search. For a lean team, that leverage is the difference between support scaling with your install base and drowning under it. See [the help center](/help-center/), or [start free](https://app.convot.io/signup) with your first app. --- # How to Add Live Chat to Your Shopify App (the Right Way) URL: https://convot.io/blog/add-live-chat-to-your-shopify-app/ Published: 2026-06-12 Summary: A practical guide to adding live chat to a Shopify app: where to put the widget, why it beats email, what to automate, how to identify merchants securely, and how to keep response times low. When a merchant gets stuck inside your app, the gap between "I have a question" and "I gave up" is measured in minutes. Live chat closes that gap. But there is a big difference between dropping a widget on your site and adding live chat that actually reduces churn and lifts your rating. Here is how to do it well. **One clarification up front, because search results blur it:** this guide is about live chat *for your Shopify app* — support for the merchants who installed it, embedded in the app's own screens inside Shopify Admin. It is not about a storefront chat widget a store owner adds to talk to shoppers. If you build a Shopify app, the support surface that moves your churn and your App Store rating lives inside the app, and that is the one worth getting right. ## Why live chat beats email for an app Email support is fine for things that can wait. The problem is that the questions that cause churn cannot wait, they happen mid-task, when a merchant is trying to do something and hits a wall. Send them to email and you have introduced a delay exactly when frustration is highest, and a chunk of them will give up or vent in a review instead. Live chat meets the question where and when it happens. It also gives you context email cannot: which screen the merchant is on, who they are, and what they are worth. For an app where setup friction is the main churn driver, chat is the highest-leverage support channel you can run. ## Put the widget where the friction is Most teams embed live chat on their marketing site and stop there. But the questions that actually cause churn happen inside the app, on the settings screen, the onboarding step, the billing page. Embed the widget on those screens, and a stuck merchant can ask without leaving the flow. Convot embeds with a single snippet on any screen of your app, with a branded widget that matches your product rather than looking like a bolted-on third party. If you are coming from another tool, the [Crisp-compatible SDK](/vs/crisp/) means migrating is mostly a copy-paste. ## Identify the merchant securely A support conversation is far more useful when you know who you are talking to. Pass the merchant's identity to the widget so every conversation is tied to the right account, and verify it with HMAC so it cannot be spoofed. Convot supports [identity verification and external_id sync](/developers/) out of the box, so the merchant's history and revenue show up automatically the moment they open a chat. ## Decide what to automate You do not need a giant chatbot. You need fast, correct answers. Start with a searchable [help center inside the widget](/help-center/) so common questions are deflected before they reach you, the full approach is in [how to reduce support tickets](/blog/self-serve-support-reduce-ticket-volume/). Then let a grounded AI agent draft replies for the routine questions, and keep a human on the conversations that genuinely need judgment. The rule: automate the repetitive, escalate the nuanced. ## Set expectations so chat does not backfire Live chat raises the expectation of an instant answer. If you cannot always be instant, be honest about it. Show clear online and away states, set business hours if you have them, and tell merchants what to expect ("typically replies within a few minutes" or "we will email you back today"). A merchant who knows you are offline and leaves a message is fine; a merchant who expected a reply and got silence is the one who churns. Expectations you set and meet beat speed you promise and miss. ## Keep response times honest The whole value of live chat is speed. Route conversations to whoever is available, put a real number on your first-response time, and hold yourself to it. A native mobile app helps, so the 2am question gets answered before it becomes a one-star review. Track median first-response time as a growth metric, not a support stat, the link between fast chat and ratings is direct, and we cover it in [customer support metrics](/blog/customer-support-metrics-shopify-apps/). ![A merchant conversation with the revenue sidebar](/shots/conversation-revenue.png) ## Make it revenue-aware Here is the part most chat tools miss: when a merchant messages you, you should know what they are worth. Convot shows each merchant's [MRR, plan, and LTV](/revenue/) beside the conversation, so you can prioritize the high-value ones instead of answering strictly first-come-first-served. A churn risk worth $300 a month should not wait behind a free-tier question, and without revenue context you cannot tell the difference. ## The takeaway Adding live chat is easy. Adding live chat that is in the right place, identified, grounded in good automation, honest about response times, and revenue-aware is what actually reduces churn and moves your App Store rating. Do that, and chat becomes the channel that catches problems before they become reviews. See [live chat for Shopify apps](/live-chat/), or [start free](https://app.convot.io/signup). --- # What to do when a merchant uninstalls your Shopify app URL: https://convot.io/blog/when-a-merchant-uninstalls-your-shopify-app/ Published: 2026-06-10 Summary: An uninstall is not the end of the conversation. Here's how to run a win-back, figure out the real reason, and turn churn into a feedback loop that fixes your product. Most app teams treat an uninstall as a closed door. The merchant is gone, the MRR is gone, and the dashboard ticks down. But the uninstall is one of the most information-rich moments in the whole relationship, if you actually use it. ## Reopen the conversation A merchant who just uninstalled has a reason fresh in their mind, and a small window where they will still talk to you. This is the moment to reach out, not with a generic "sorry to see you go" survey, but as a continuation of whatever support thread you already had with them. Convot does this automatically: when a merchant uninstalls, it reopens the support conversation so you can ask what happened and, when it makes sense, offer a save, an annual discount, a fixed bug, a walkthrough of the feature they missed. ## Separate the saveable from the lost Not every uninstall is worth chasing. A merchant who closed their store is gone. A merchant who left over a pricing objection or a confusing setup step might come back with one good reply. The trick is knowing which is which, fast, so you spend your energy on the saveable ones. Pair the win-back with [revenue context](/revenue/): a merchant who was paying $150 a month is worth a personal message; a free-tier tester who never activated probably is not. ## Find the real reason Exit surveys lie. Merchants pick whatever option closes the dialog fastest. The honest signal is in the support history, the questions they asked, the bug they hit, the feature they requested. Convot's [AI churn attribution](/ai/) reads that history after an uninstall and tells you whether support was the cause, then rolls it into a support-attributable churn percentage. Instead of guessing, you get a number that says how much of your churn you can actually fix with better support, and which issues to fix first. ## Close the loop The point of all this is not just to save the occasional merchant. It is to turn churn into a product feedback loop. Every uninstall that traces back to the same missing feature or the same confusing step is a prioritized to-do. Fix the top one, watch that slice of churn shrink, repeat. An uninstall is not a closed door. It is the most honest feedback you will get all month. Use it. See how [churn-save and attribution](/revenue/) work, or [start free](https://app.convot.io/signup). --- # Reading your Shopify Partner revenue: MRR, ARR, and LTV explained URL: https://convot.io/blog/shopify-partner-revenue-mrr-arr-ltv/ Published: 2026-06-08 Summary: MRR, ARR, ARPU, LTV: the numbers behind a Shopify app business, what each one actually means, and how to read them from the Partner API without a spreadsheet. If you build a Shopify app, your business is a subscription business, whether you think of it that way or not. And subscription businesses are run on a small set of numbers. Here is what they mean and how to actually read them from your Shopify Partner data. ## MRR: monthly recurring revenue MRR is the predictable revenue you collect every month from active subscriptions. It is the single most important number for an app business because it is recurring: a merchant on a $29 plan adds $29 to MRR every month until they leave. Watch the trend, not the snapshot. MRR going up and to the right means your acquisition is outpacing your churn. ## ARR and ARPU ARR is simply MRR times twelve, the annualized run rate. It is useful for talking about the size of the business. ARPU, average revenue per user, is MRR divided by active merchants. Rising ARPU means you are moving merchants up your plans or attracting higher-value ones. ## LTV: lifetime value LTV estimates how much a merchant is worth across their whole relationship with you. A rough version is ARPU divided by your monthly churn rate. If your average merchant pays $30 a month and 4% of merchants churn monthly, the average lifetime is about 25 months and LTV is roughly $750. LTV is what tells you how much you can afford to spend acquiring a merchant. ## Reading it without a spreadsheet The Shopify Partner API exposes the raw transactions and lifecycle events behind all of this. The hard part is turning them into a live dashboard. Most app teams either build one or pay for a tool like [ChartMogul or Baremetrics](/vs/chartmogul/). Convot does it for you: connect your Partner account once and it syncs gross MRR, net payout, ARR, ARPU, LTV, installs, and uninstalls into a [revenue dashboard](/revenue/), and then puts each merchant's MRR, plan, and LTV beside the support conversation. The number you need to triage a ticket is right where you are already working, and you skip the separate analytics subscription. ![The merchant's MRR, plan, and LTV beside a live conversation](/shots/conversation-revenue.png) ## Why it belongs in support Metrics in a dashboard you check weekly are useful. The same metrics beside a live conversation are powerful. When you can see that the merchant messaging you is worth $1,800 in lifetime value, you answer differently. That is the whole idea behind [revenue-aware support](/revenue/). [Start free](https://app.convot.io/signup) and connect your Partner account in minutes. --- # Meet Cove AI: a support agent that answers from your docs, not its imagination URL: https://convot.io/blog/meet-cove-ai-grounded-support-agent/ Published: 2026-06-05 Summary: Cove AI answers your merchants grounded only in your own knowledge, cites its sources, and escalates when it isn't sure. Here's how it works, and why a grounded agent beats a generic chatbot for Shopify apps. Most AI support chatbots are optimized for one thing: always having an answer. For a Shopify app, that's exactly the wrong instinct. A confidently wrong answer about a setting, a plan, or a billing question doesn't deflect a ticket, it creates one, and sometimes a one-star review along with it. If you want to understand why this happens and how grounding fixes it, [What is an AI support agent? (And how to keep it from hallucinating)](/blog/what-is-an-ai-support-agent/) covers the mechanics. So we built Cove AI the other way around. It answers from your knowledge and only your knowledge, it tells you where each answer came from, and when it isn't sure, it hands off to a human instead of guessing. ## Grounded in your docs, by design Cove AI trains on four sources you control: - **Help-center articles** - one click embeds your existing published articles, no manual setup. - **A website crawl** - point it at a docs or help site and it learns from the pages. - **Uploaded files** - drop in manuals or FAQs (TXT, MD, CSV). - **Hand-written Q&A** - exact answers to common questions, grouped by topic. Under the hood, every question runs through hybrid retrieval (semantic search plus keyword search) and a cross-encoder reranker that picks the passages most relevant to what was actually asked. Cove AI answers from those passages and cites them back to the customer. If the answer isn't in your sources, it won't invent a price, a policy, or a step. No hallucinated answers under your brand. ## It knows when to stop A confidence threshold decides when Cove AI answers versus when it escalates. Below the line, it abstains and routes the conversation to your team, with a warm, context-specific handoff line so the customer gets a real transfer, not a cold dead end. You can also fence off topics it must always send to a human, billing disputes, cancellations, anything legal, and Cove AI never touches them. Customer messages are treated as untrusted input, so attempts to jailbreak the agent or extract its instructions get politely declined. ## Start in Suggest. Graduate to Auto. You don't have to trust an AI agent on day one. Cove AI starts in **Suggest** mode: it drafts replies your agents approve, edit, or reject in one click. The analytics track how often your team accepts those drafts, and tell you when the acceptance rate is high enough that Cove AI is ready to run on its own. When it is, you have options: - **Auto** - Cove AI replies to customers directly. - **Auto when offline** - it only replies off-hours, covering nights and weekends while your team owns the day. - **Hybrid** - suggest while you're online, auto when you're not. It graduates at the pace your team actually trusts it. ## Plan-aware, because you build the app This is the part generic chatbots can't do. Connect your Shopify Partner account and Cove AI answers with each customer's billing reality in mind: whether they're on a free install or paying, their MRR, and whether they're Shopify Plus. So when a free-install merchant asks about a paid feature, Cove AI is honest that it's a paid feature instead of pretending it doesn't exist, and it can nudge an upgrade when you want it to. It only ever cites prices and links that appear in your own sources. ## See what it deflects, and what it's missing Cove AI ships with its own analytics: deflection rate, conversations handled, CSAT, and cost over time. And **Knowledge Gaps** clusters the recurring questions Cove AI couldn't answer, so instead of guessing, you know exactly which Q&A or article to add next to deflect more. ## Simple, usage-based pricing Cove AI is an add-on you can switch on for any plan, priced at **$0.20 per resolved conversation**. You pay when it actually does the work, not per seat, and the in-product cost tracking means you're never surprised by a bill. [See how Cove AI works →](/cove-ai/) --- # How to Reduce Customer Churn: A Retention Playbook URL: https://convot.io/blog/reduce-shopify-app-churn/ Published: 2026-06-03 Summary: A practical playbook to reduce customer churn: how to measure it, spot at-risk customers early with health scores, save the ones worth saving, recover failed payments, and grow net revenue retention. Every signup is exciting. Every cancellation is a number you try not to look at. But churn is the difference between a business that compounds and one that runs on a treadmill, paying to acquire customers who quietly leave a few months later. At 5 percent monthly churn you lose nearly half your customers a year, and growth has to outrun that leak before it does anything else. The good news: most churn is not random. It clusters around a handful of moments you can actually influence. Here is how to measure it, see it coming, and bring it down. ## First, measure churn properly You cannot reduce what you do not measure, and a single "churn rate" hides two very different problems. - **Customer churn rate**: cancelled customers in a period divided by customers at the start. The count of logos leaving. - **Revenue churn rate**: lost recurring revenue divided by revenue at the start. What actually hits the bank. The two diverge in a way that matters. If you lose ten free or low-value customers but keep your big accounts, customer churn looks scary while revenue churn barely moves. Watch both, and obsess over revenue churn. The best teams also track **net revenue retention** (more on that below), which folds in upgrades and can offset churn entirely. ## See it coming: leading indicators and a simple health score By the time someone cancels, you have already lost. The leverage is in the weeks before, and the signals are there if you look. The strongest leading indicators of churn are almost always **declining usage** (logins, core actions, active seats), **stalled adoption** (never reached the feature that delivers the value), and **negative support sentiment** (slow resolutions, repeated issues, frustration in the thread). You do not need a data team to act on this. Build a simple health score: pick three or four signals, score each green/yellow/red, and review the red and yellow accounts weekly. A customer whose usage halved and who has an unresolved ticket is telling you they are leaving. Reaching out before they decide is far cheaper than a win-back after. ## Churn starts long before the cancellation The decision is usually weeks old. It was made the night they hit a confusing setup step and got no answer, the week a bug went unfixed, or the moment a competitor shipped the feature they had asked for. So the first move is not a win-back email. It is faster, better support, and a way to see which conversations are turning sour while they are still fixable. Convot's [frustration escalation](/ai/) reads every conversation and flags the ones turning hostile in real time, so you can step in before a quiet customer becomes a one-star review and then a cancellation. The same fast support that heads off the cancellation is what [earns more 5-star reviews](/blog/how-to-get-more-shopify-app-reviews/) too: the merchant you rescue is often the one who recommends you. ## Prioritize by revenue: know who is worth saving Not every customer deserves the same response time. A customer paying you $200 a month who hits a bug is a different priority from a free-tier user kicking the tires. Most support tools show you none of that, so your team treats a churn risk worth thousands the same as a tire-kicker. Convot shows each customer's [MRR, plan, and lifetime value](/revenue/) right beside the conversation, so you protect the revenue that matters first. It is the same idea from [customer support metrics](/blog/customer-support-metrics-shopify-apps/): connect support activity to revenue, and your priorities sort themselves. ![Each customer's MRR, plan, and LTV shown beside the support conversation](/shots/conversation-revenue.png) ## Win the first two weeks A huge share of churn happens early, before a customer ever becomes a habit. They signed up, got stuck, and drifted. The fix is proactive onboarding: reach out during the first session, answer the first question fast, and get them to the moment the product actually pays off. Track where new customers stall, that cluster is both your biggest support generator and your biggest early-churn driver. ## Reduce the friction that pushes people out High effort is one of the most reliable predictors of churn. Every time a customer waits, repeats themselves, or hunts for an answer, you nudge them toward the exit. A strong self-serve layer, a searchable help center and clear docs, removes that friction at the moment it appears. The [ticket deflection playbook](/blog/self-serve-support-reduce-ticket-volume/) covers how to build it, and the payoff is not just fewer tickets, it is lower churn. ## Recover involuntary churn Some of your churn is not a decision at all. It is failed payments: expired cards, insufficient funds, and bank declines. For many subscription businesses this is a large and entirely recoverable slice of churn. Set up dunning, retry failed charges on a smart schedule, and email customers before and after a card expires. Fixing involuntary churn is often the single fastest retention win available, because nobody actually wanted to leave. ## Segment why people leave, and fix the biggest bucket "Reduce churn" is too vague to act on. Make it specific by tagging every cancellation with a reason: price, missing feature, poor onboarding, an unresolved bug, switched to a competitor, or no longer needed. After a few weeks the buckets tell you where to spend. Churn driven by onboarding is a product fix; churn driven by slow support is a staffing or tooling fix; churn driven by a missing feature is a roadmap call. Without segmentation you treat all churn as one problem and fix none of it well. ## The math that compounds: net revenue retention The most powerful retention metric is **net revenue retention (NRR)**: the revenue you keep from existing customers, including upgrades, minus downgrades and cancellations. When expansion from growing accounts outweighs the revenue you lose, NRR climbs above 100 percent and your existing base grows even with zero new customers. That is the holy grail, negative churn. You get there by making successful customers more valuable over time (usage-based growth, upsells, new features they adopt), not just by plugging leaks. Retention and expansion are the same muscle: a customer who is succeeding renews and expands; one who is struggling churns. ## Should you run save offers? When a customer hits cancel, a pause option or a targeted discount can recover some who are leaving for fixable reasons (temporary budget, a feature shipping soon). But save offers mask the real problem if you lean on them. A discount does not fix bad onboarding or slow support, it just delays the churn and trains customers to threaten leaving. Use them sparingly, and always capture the reason so the save offer is data, not a band-aid. ## Learn from every cancellation You will never save everyone. But you should always know why you lost the ones you did. When a customer cancels, Convot reopens the conversation and its [AI](/ai/) tells you whether support was the cause, then rolls it up into a support-attributable churn percentage on your dashboard. That number turns a vague fear into a specific to-do. It also replaces a separate revenue dashboard, so you are not paying for a [Mantle or ChartMogul](/vs/mantle/) subscription on the side. ## The short version Reduce churn by doing it in order: measure customer and revenue churn separately, build a health score to catch risk early, prioritize by revenue, win the first two weeks, remove friction with self-serve, recover failed payments, segment why people leave, grow net revenue retention through expansion, and learn from every cancellation. Do that consistently and churn stops being a tax and becomes a signal that tells you exactly what to build and fix next. See how [revenue-aware support](/revenue/) works, or [start free](https://app.convot.io/signup) and put revenue context beside every conversation. --- # Customer Support Metrics & KPIs: The Ones That Actually Matter URL: https://convot.io/blog/customer-support-metrics-shopify-apps/ Published: 2026-05-22 Summary: A practical guide to customer support metrics and KPIs: how to calculate each, what good looks like, and which few actually predict retention, reviews, and revenue. Most support tools will happily show you twenty charts. For a lean team, that is the problem, not the solution. A wall of widgets feels like insight but rarely changes what you do on Monday morning. The teams that actually improve track a handful of customer support metrics, know what "good" looks like for each, and act on the trend. This guide covers the customer support metrics and KPIs worth knowing, how to calculate each, what a healthy benchmark looks like, and which few genuinely predict retention and revenue. Treat the benchmarks as directional: the right target depends on your channel, your customers, and your price point. ## Speed metrics: how fast you respond and resolve ### First response time (FRT) The time from a customer's first message to your first human reply. It is the metric customers feel most, because waiting in silence is what turns a small question into frustration. Track the **median**, not the average. One ugly weekend ticket can drag an average into meaninglessness while the median tells you what a typical customer actually experiences. Good looks like under a minute for live chat during active hours, and under a few hours for email. The exact number matters less than the consistency. ### Average resolution time The time from first message to "resolved." First response time gets attention; resolution time delivers the outcome. Watch the long tail here, not just the average. The handful of tickets that drag on for days are usually the ones that become bad reviews or churned customers, so segment them out and ask why they stalled. ## Quality metrics: how well you solve problems ### Customer satisfaction (CSAT) A direct "was this helpful?" or 1 to 5 rating after a resolution. CSAT is your cleanest quality signal. Calculate it as the percentage of positive responses out of all responses. Above 90 percent is strong for most software teams. At low volume it is noisy, so watch the trend across weeks rather than obsessing over a single bad day. ### First contact resolution (FCR) The share of issues solved in a single interaction, with no back and forth and no reopen. High FCR is the quiet driver of both satisfaction and low cost: every reopened ticket is a customer re-explaining themselves and an agent re-loading context. Around 70 to 75 percent is a healthy target. If FCR is low, the fix is usually better macros, clearer docs, or surfacing more context to the agent up front. ### Customer effort score (CES) How hard the customer had to work to get helped, usually a "this was easy" agree/disagree prompt. Effort predicts loyalty better than delight does: customers rarely reward you for low effort, but high effort reliably pushes them toward the exit. Lower is better. ## Volume and efficiency metrics: can you scale? ### Ticket volume The raw count of incoming conversations. On its own it means little. More tickets is not inherently bad, slow tickets are. Volume becomes useful when you normalize it (tickets per customer, per account, or per 100 active users) and watch the direction. A spike in tickets-per-customer after a release is a product signal, not a staffing problem. ### Ticket deflection rate The share of customers who get an answer without ever opening a conversation, usually through a help center or in-product content. Deflection is what lets support volume grow slower than your customer base. If conversations rise while your team stays flat, deflection is doing the work. We go deep on this in [how to reduce support tickets](/blog/self-serve-support-reduce-ticket-volume/). ### Backlog and average ticket age How many conversations are open and how old they are. A growing backlog is the earliest warning that demand is outrunning capacity, well before it shows up in CSAT. Watch the age of the oldest open tickets, because those are where quiet frustration compounds. ## Loyalty: the long-game metric ### Net promoter score (NPS) "How likely are you to recommend us?" on a 0 to 10 scale, scored as the percentage of promoters minus detractors. NPS is a relationship metric, not a ticket metric, so survey periodically rather than after every interaction. It is most useful as a slow-moving trend line and as a prompt: read the comments behind the scores, that is where the actual product and support insight lives. ## How to choose which KPIs to track You do not need all of these. Tracking twelve metrics usually means acting on none. Pick three or four that map to your current bottleneck: - **Customers complaining you are slow?** First response time and backlog. - **Solving things but customers still churn?** FCR, CES, and resolution time. - **Drowning as you grow?** Deflection rate and tickets-per-customer. Choose, watch the trend, and move the number. Then revisit the set as your constraint changes. ## The metric most teams miss: support's effect on revenue Here is the one that almost no support dashboard shows, and the one that matters most for a software business: **what support is doing to retention and revenue.** For a broader view of how to build the support function around these outcomes, [SaaS Customer Support: A Practical Guide for Lean Teams](/blog/saas-customer-support/) covers the full picture. A fast, high-quality support interaction is not just a closed ticket. It is a renewal you kept, a downgrade you prevented, a five-star review you earned. The reverse is also true: a slow or unresolved issue is often the quiet first step toward a cancellation. If you can connect support activity to revenue outcomes, support stops being a cost center on a spreadsheet and becomes a retention lever you can actually manage. For SaaS and Shopify app teams specifically, the highest-leverage move is to put each customer's revenue context, their plan, monthly value, and churn risk, beside the conversation, so you can see which slow tickets are actually expensive. Then close the loop: when a customer does leave, look back at whether support was the cause. That feedback is what tells you which metric to fix next. We cover the playbook in [how to reduce customer churn](/blog/reduce-shopify-app-churn/), and the link between fast support and ratings in [how to get more reviews](/blog/how-to-get-more-shopify-app-reviews/). ## How to track these without building reports You should not have to wire up a data pipeline to know your first response time. [Convot](/) tracks first response and resolution time automatically, groups activity by customer so you see a complete history in one place, and shows [live revenue intelligence](/revenue/), plan, value, and churn risk, beside every conversation. Its [AI](/ai/) can also attribute churn back to support after a customer leaves, which turns your metrics from a scoreboard into a to-do list. > Pick the few metrics that predict retention, watch the trend, and improve the worst one. First response time is almost always the right place to start. ## The takeaway Customer support metrics are only useful when they change what you do. Skip the vanity dashboard. Track speed (first response, resolution), quality (CSAT, FCR), and efficiency (deflection, backlog), pick the three or four that map to your current bottleneck, and connect them to the outcome that pays the bills: retention. Answer fast, resolve completely, deflect the easy questions, and watch the numbers that tell you whether you are doing exactly that. [Try Convot free](https://app.convot.io/signup) and get first response and resolution tracking, with revenue context, out of the box. --- # Why we built Convot for Shopify app developers URL: https://convot.io/blog/why-convot-for-shopify-app-developers/ Published: 2026-05-21 Summary: Shopify apps live and die by their App Store rating, and that rating is won or lost in support. Here's why we built a chat platform specifically for app teams. If you build a Shopify app, you already know the uncomfortable truth: your App Store rating is your growth engine. It decides your ranking, your install rate, and whether a merchant trusts you enough to connect their store. And nothing moves that rating faster, in either direction, than support. A merchant hits a confusing setup step at 11pm their time. They send a message. If they hear back in two minutes, you've earned a five-star review. If they wait two days, you've earned a one-star review that quietly taxes every future install. Every existing tool is built for someone else. ## The tools weren't built for us Gorgias is built for Shopify *merchants*, the stores selling t-shirts and supplements. Intercom is built for enterprise. Crisp and the rest are general-purpose. None of them are built for the people building *on* Shopify: small, technical teams who ship an app and then have to support thousands of non-technical merchants across the world. If you're evaluating Crisp specifically, [The best Crisp alternatives for Shopify app teams (2026)](/blog/best-crisp-alternatives-shopify-app-teams/) is a direct comparison. That mismatch shows up everywhere: - **Language.** Your merchants are global. A generic inbox makes you copy-paste into Google Translate. - **Context.** You want to see the store, the plan, the install date, not a blank contact card. - **Integration.** You think in webhooks and APIs. Most tools think in "book a demo." ## What Convot does differently We built Convot around three bets: 1. **Speed is the product.** A native mobile app with push notifications means you answer that 2am question before it becomes a review, not after. That response speed is exactly what drives [how to get more 5-star reviews for your Shopify app](/blog/how-to-get-more-shopify-app-reviews/). 2. **Language shouldn't be a tax.** Claude-powered live translation detects your merchant's language and translates both ways, automatically, in 25+ languages. 3. **It should feel like your stack.** Signed webhooks fire on nine events, a REST API exposes your data, and `external_id` keeps your own customer records in sync. No sales call to integrate. > Support isn't a cost center for a Shopify app. It's the difference between a 4.2 and a 4.9. That's also why it pays to [catch frustrated merchants before they leave a 1-star review](/blog/catch-frustrated-merchants-before-bad-reviews/). ## Coming from Crisp? You can import your conversations, contacts, and help-center articles in one click. Keep your history, keep your merchants, and lose the parts that slowed you down. We're just getting started. If you build Shopify apps, [try Convot free](https://app.convot.io/signup), your first app is on us. See [our pricing](/pricing/) for a full breakdown. --- # How to support Shopify app merchants in any language URL: https://convot.io/blog/multilingual-support-shopify-apps/ Published: 2026-05-19 Summary: Shopify merchants are global, but most app teams only support English. Here's how to offer multilingual customer support without hiring a multilingual team. Shopify powers merchants in over 175 countries. If your app is on the App Store, a meaningful share of the people installing it don't speak English as a first language, and many don't speak it at all. Yet most app teams support only in English. The result is quiet churn: a merchant in São Paulo or Berlin hits a problem, can't explain it well in English, gives up, and uninstalls (or leaves a frustrated review). You don't need to hire a multilingual support team to fix this. Here's how small app teams handle global support. ## Why language is a bigger deal than it looks - **Trust.** Merchants are connecting their store and customer data to your app. Getting help in their own language signals that you take them seriously. - **Resolution quality.** A merchant describing a bug in their native language gives you far better detail than one fighting through a second language. - **Reviews.** "They even helped me in Portuguese" is the kind of detail that ends up in a five-star review. ## Option 1: Human translation (doesn't scale) You can copy-paste into a translation tool for each message. It works for a handful of conversations, but it's slow, error-prone, and breaks down the moment volume grows. It also adds latency to exactly the metric that matters most, [first-response time](/blog/customer-support-metrics-shopify-apps/). ## Option 2: Pre-translated help articles A multilingual help center deflects common questions before they reach you. It's worth doing, but it only covers the predictable questions, not the live back-and-forth of real support. ## Option 3: Automatic live translation The scalable answer is translation built into the inbox. The merchant writes in their language, you read it in yours, you reply in yours, and they receive it in theirs, automatically. This is how [Convot](/) handles it. Our inbox uses Claude-powered live translation that: - **Auto-detects** the merchant's language on their first message. - **Translates both directions**, their message into your language, your reply into theirs. - **Covers 25+ languages** with zero configuration. Your team keeps working in English. The merchant feels supported in their own language. Nobody copy-pastes anything. > The goal isn't to speak every language. It's to make every merchant feel like you do. ## Pair it with a multilingual help center Live translation handles conversations; a translated help center handles self-serve. Convot supports help-center articles in multiple languages on your own domain, searchable right inside the chat widget, so merchants get answers in their language before they ever open a conversation. ## The takeaway Global merchants are already installing your app. Supporting them in their language is one of the cheapest ways to cut churn and lift your rating, especially once it's automatic. [Try Convot free](https://app.convot.io/signup) and reply to any merchant, in any language, from day one. --- # The best Crisp alternatives for Shopify app teams (2026) URL: https://convot.io/blog/best-crisp-alternatives-shopify-app-teams/ Published: 2026-05-15 Summary: Outgrowing Crisp? Here's an honest comparison of the best Crisp alternatives for Shopify app developers, Intercom, Tidio, Chatwoot, and Convot, and how to choose. Crisp is a solid starting point for live chat. But as a Shopify app grows, teams often run into the same walls: pricing that climbs with contacts, a generic inbox that isn't built for app-merchant support, and a migration that feels scary enough to keep you stuck. If you're evaluating a switch, here's an honest look at the main options in 2026, including where we think each one fits. ## What to look for as a Shopify app team Before the list, the criteria that actually matter for app developers (not merchants): - **Speed to respond**, live chat plus a real mobile app, so you answer before a question becomes a one-star review. - **Language coverage**, your merchants are global; translation shouldn't be a manual copy-paste. - **A self-serve help center** to deflect repetitive questions. - **Developer surface**, webhooks and an API to wire support into your stack. - **A painless migration** off your current tool. ## Intercom The enterprise standard. Deep automation, a strong AI agent, and a huge feature set. The trade-offs are price and complexity, it's built for large support orgs, and the per-seat-plus-usage pricing adds up fast for a small app team. Great if you're scaling support headcount; heavy if you're a two-person team. ## Tidio Friendly, affordable, and popular with smaller e-commerce businesses. Good live chat and a capable AI agent. It's aimed more at merchants than at the developers building apps, so some of the workflow is e-commerce-storefront-shaped rather than app-support-shaped. ## Chatwoot Open-source and developer-friendly, with a self-hostable option that appeals to technical teams who want control and no per-seat pricing. The trade-off is that you run and maintain it yourself (or pay for their cloud), and you assemble some of the polish. ## Convot We built [Convot](/) specifically for the teams **building** Shopify apps. It's live chat, an embedded help center, and a native iOS/Android app, with a few things we found app teams need: - **Claude-powered live translation** in 25+ languages, both directions, automatically, so you support merchants worldwide without a translation tool. - **A native mobile app with push** so you can answer that 2am merchant before it becomes a review. - **9 signed webhooks and a REST API** to hook support into your own systems. - **One-click Crisp import**, bring your conversations, contacts, and help-center articles over without losing history. It's intentionally focused: not an enterprise suite, but the support stack a Shopify app team actually uses. ## How to choose | If you... | Consider | |---|---| | Are scaling a large support org | Intercom | | Want a friendly, low-cost merchant tool | Tidio | | Want open-source / self-hosting | Chatwoot | | Are a Shopify app team that lives by its rating | Convot | ## Switching is the easy part The biggest reason teams stay on a tool they've outgrown is migration fear. With Convot you can [import everything from Crisp in one click](/blog/how-to-get-more-shopify-app-reviews/), conversations, contacts, and articles, and keep your history. If you build Shopify apps, [try Convot free](https://app.convot.io/signup). Your first app is on us, no credit card required. --- # How to Get More 5-Star Reviews for Your Shopify App URL: https://convot.io/blog/how-to-get-more-shopify-app-reviews/ Published: 2026-05-12 Summary: Your Shopify App Store rating drives installs, ranking, and trust. A practical playbook for earning more 5-star reviews, with the timing, the ask, and why support is the lever that moves it. For a Shopify app, your App Store rating is not a vanity metric. It decides where you rank in App Store search, how many of your listing visitors actually install, and whether a merchant trusts you enough to connect their store. A drop from 4.8 to 4.4 quietly taxes every future install, and because ratings are an average, early reviews follow you for a long time. The good news: reviews are earnable, and you do not earn them by nagging harder. You earn them by making merchants successful and then asking at the right moment. Here is a playbook that works for small app teams. ## 1. Ask at the moment of value, not at install The worst time to ask for a review is right after install. The merchant has not gotten value yet, so the best you can hope for is indifference. The best time is right after a win: their first successful sync, their first sale through your feature, a milestone hit, or a support issue you just resolved fast. Map your app's "aha moment", the point where the merchant first feels the value, and trigger the review prompt there. A merchant who just felt the payoff is far more likely to leave five stars than one you interrupted mid-setup. ## 2. Make support your review engine This is the lever almost everyone underuses. Most one-star reviews are not really about your product. They are about a question that went unanswered for two days. A merchant hits a confusing step, sends a message, hears nothing, and vents in public where it costs you future installs. Flip it. When you answer in minutes instead of days, the same merchant who would have left one star often becomes a promoter. Fast, human support is the single biggest lever on your rating. > Every unanswered merchant is a one-star review in progress. Every fast, helpful reply is a five-star review waiting to happen. A few support habits that protect your rating: - **Respond fast, even if it is just "looking into this."** Acknowledgement buys time and goodwill. - **Be reachable where merchants already are.** Live chat on your app pages beats a buried support email. - **Do not let language be a barrier.** A Spanish- or German-speaking merchant who gets help in their own language remembers it. Convot uses [live translation](/blog/multilingual-support-shopify-apps/) so you can reply in 25+ languages automatically. ## 3. Turn resolved tickets into reviews The best review ask is the one that follows a save. When you have just turned a frustrated merchant into a happy one, that is the highest-converting moment to ask, and it is completely organic. Build the habit: after a fast, successful resolution, a simple "glad that is sorted, if the app has been useful, a quick App Store review really helps us" converts far better than a generic in-app banner. You are not buying a review; you earned it, and you are asking at the peak of goodwill. ## 4. Get the ask mechanics right How you ask matters as much as when: - **Make it one tap.** Link straight to your App Store review page; every extra step loses people. - **Personalize the trigger, not just the copy.** Tie the prompt to a real event in their account, not a fixed day count. - **Never incentivize or buy reviews.** Offering discounts or rewards for reviews violates Shopify's policy and risks your listing. Earn them instead. - **Do not over-ask.** One well-timed prompt beats three poorly-timed ones that train merchants to dismiss you. ## 5. Respond to every review, especially the bad ones Shopify lets you reply publicly, and future merchants read those replies. A calm, helpful response to a one-star review often does more for trust than the review itself does damage, because it shows you care and you ship fixes. When you resolve the underlying issue, politely ask the reviewer if they would consider updating their rating. Many will. And if a review is abusive or clearly violates Shopify's review policy, you can report it for removal, but lead with a genuine reply first; that is what other merchants see. ## 6. Reduce the tickets that cause frustration Every confusing step is a future support ticket and a potential bad review. Cut them at the source with a [help center inside your support widget](/blog/self-serve-support-reduce-ticket-volume/) so merchants self-serve the easy answers, and proactive messages at known friction points ("Need a hand connecting your store?"). Deflecting the easy questions frees you to answer the hard ones fast, which is exactly what protects the rating. ## 7. Track first-response time like a growth metric If you measure one support number, measure median first-response time. It correlates with ratings more tightly than almost anything else, because speed is what separates a vented one-star from a grateful five-star. We go deeper in [the customer support metrics that matter](/blog/customer-support-metrics-shopify-apps/). ## Common questions about Shopify app reviews **How many reviews does a Shopify app need?** There is no magic number, but volume and recency both matter for ranking and trust. A steady trickle of recent 5-star reviews signals an actively maintained app; a wall of two-year-old reviews looks stale at the same average. Aim for a consistent flow tied to real merchant wins, not a one-time push. **How do I ask for a review without breaking Shopify's policy?** Ask, but never incentivize. You can prompt a merchant to review after a genuine win; you cannot offer a discount, credit, or reward in exchange, which violates Shopify's review policy and can risk your listing. Tie the ask to a real moment of value and make it one tap to your App Store page. **What is the fastest way to raise a low rating?** Fix the support gap first. Most one-star reviews trace back to a slow or missing reply, so a faster first-response time, plus a calm public reply to your existing negative reviews with a real fix, moves the average faster than chasing new reviews alone. ## The takeaway You do not earn five-star reviews by asking harder. You earn them by making merchants successful, answering them fast and in their language, and asking at the moment they just felt the value. Support is where that happens, so treat it as the growth channel it is. That is exactly why we built [Convot for Shopify app teams](/): [live chat](/live-chat/), a help center, live translation, and a native mobile app so you never miss the message that becomes a review. [Start free](https://app.convot.io/signup), your first app is on us.