That distinction matters more than any feature grid, because per-seat and per-resolution billing punish two completely different things. Below is how to tell which one is actually hurting you, followed by an honest walk through the realistic options and the questions worth asking before you migrate anything.
Why does Intercom's pricing feel wrong for a small team?
Intercom is a capable product. It became the default for venture-backed SaaS companies, and the pricing model reflects that origin: you pay per seat, and on newer plans you also pay per AI resolution. For a company with a dedicated support department and a finance team that models headcount a year out, that is a defensible way to buy software. For a nine-person business where the founder, the bookkeeper, two technicians and a weekend contractor all occasionally need to see a customer thread, it creates a specific kind of friction.
The friction is not only the invoice. It is that you start rationing access. The office manager who could answer half the inbound questions does not get a login, because adding her costs money every month forever. The part-time weekend cover person shares a password with someone else. The owner stops looking at conversations because her seat is the one that got cut. None of that shows up as a line item, but it is the real cost: your support quality is now shaped by your billing model instead of by who is best placed to help.
Per-resolution AI pricing adds a second problem. When each automated answer carries a fee, you become quietly reluctant to let the AI handle more. A busy month turns into a bigger bill, which is the opposite of what you want from automation. Growth should reduce your cost per conversation, not raise your total unpredictably.
So before you shop, name the problem precisely. If you have few people and enormous volume, per-seat pricing is not your enemy and some cheap per-seat tools will serve you well. If you have many occasional helpers, per-seat pricing is exactly the thing to escape.
What does flat-rate pricing actually change day to day?
Flat pricing changes who you invite. That sounds small and it is not.
When seats are free, the natural move is to put everyone who touches customers into the same inbox. Sales adds context on a thread before support replies. The technician who did yesterday's install answers the follow-up himself rather than relaying it. The owner reads the week's conversations on Sunday evening without anyone having to deprovision a login to make room. Internal notes replace forwarded email chains, and the customer's history lives in one place instead of three inboxes and a phone.
The second change is that you can stop forecasting. A flat monthly figure is a number you can put in a budget and forget, which for most small businesses is worth more than squeezing out a slightly lower theoretical minimum.
Corebee is built on that model: $99 per month flat, with unlimited teammates and unlimited AI conversations, no per-seat and no per-resolution fees. Its shared inbox covers web chat, email, voice through an AI receptionist that tells callers it is an AI, WhatsApp, Slack, Facebook, Instagram, SMS and Telegram, and it answers from your website and knowledge base, handing off to a human when the customer asks or when it is unsure. It also books meetings through Cal.com or Calendly and syncs contacts to HubSpot, Salesforce, Pipedrive, monday and Zoho. Setup takes around ten minutes and needs no developer. If you want the direct feature-by-feature framing against the incumbent, there is a dedicated Zendesk alternatives breakdown that applies the same logic.
To be fair about the trade-off: flat pricing is not automatically cheaper. A solo operator who needs one seat and handles a handful of chats a week will find cheaper entry points elsewhere. Flat rate wins when the headcount that needs access is larger than the headcount you would willingly pay for.
Which Intercom alternatives are worth a serious look?
Help Scout is the most natural landing spot for teams whose work is mostly email and who care about tone. The interface is calm, the shared inbox model is mature, and it does not try to be a marketing platform at the same time. It is per-seat, so it solves the "Intercom is heavy and expensive" problem better than it solves the "too many people need access" problem.
Freshdesk sits at the opposite end: a full service desk with ticketing, SLAs, automation rules and a free tier for small agent counts. If your support involves queues, escalation paths and reporting that someone actually reads, Freshdesk gives you the structure. The cost is complexity; the setup is a project rather than an afternoon.
Chatwoot is the open-source option. Self-host it and the per-seat question disappears entirely, because you are paying for a server rather than for people. That is genuinely the cheapest path at scale, and it is the right answer if you have someone comfortable running a service: upgrades, backups, uptime and the AI configuration all become your responsibility. Chatwoot also offers a hosted plan if you want the product without the operations.
Tidio is aimed squarely at small ecommerce and local businesses that mainly want live chat plus a bot on the website. It is quick to install and inexpensive to start. Its scope is narrower than Intercom's, which is a feature if website chat is most of your volume and a limitation if you need the phone and a real ticketing backbone.
Gorgias is the specialist choice for Shopify and ecommerce more broadly. Its deep commerce integrations mean agents can see and act on orders, refunds and subscriptions from inside the conversation. If you are not selling physical products online, that advantage is mostly irrelevant to you.
Zendesk deserves a mention even though it is usually the thing people are also escaping. It is the most complete and most configurable of the lot, with the largest app ecosystem. For a small team it is typically more platform than the problem requires, and the pricing model shares Intercom's per-seat shape.
So the shortlist by situation: many occasional helpers and unpredictable AI volume points to flat rate; email-first and quality-conscious points to Help Scout; process and queues point to Freshdesk; technical and cost-obsessed points to self-hosted Chatwoot; website chat only points to Tidio; Shopify points to Gorgias.
How should you compare the real cost rather than the sticker price?
List every person who would benefit from access, not every person you currently pay for. Include the part-timer, the contractor, the bookkeeper who handles billing questions and the owner. That list is your true seat count, and comparing per-seat quotes against the count you have been rationing yourself down to is how teams talk themselves into the wrong tool.
Next, look at how the vendor charges for automation. Per-resolution fees mean a busy month is an expensive month, so the bill grows with exactly the thing you were trying to make cheaper. Unmetered AI conversations make that variable disappear. Either model can be fine, but you should know which one you are buying.
Then count channels against what customers actually use. If a meaningful share of your inbound is phone calls or WhatsApp, a tool that covers web chat and email well is going to leak work into personal phones and private accounts, which is where context and accountability go missing.
Finally, price the migration honestly in hours. Who writes the knowledge base articles, who reviews the first week of AI answers, who connects the CRM, and who tells customers the chat widget has changed? A tool you can configure in an afternoon and a tool that needs a two-week implementation project have very different real costs even when the subscriptions match.
What should you watch out for when switching away from Intercom?
The first risk is losing history. Export your conversations before you cancel anything, and keep the archive somewhere searchable even if the new tool cannot import it cleanly. Customers reference past threads, and "I cannot see what we agreed last quarter" is a bad start with a new system.
The second risk is treating the AI as a switch you flip rather than something you train. Whatever vendor you pick, the quality of automated answers tracks the quality of what you feed it. Thin, outdated or contradictory help content produces thin, confident and wrong replies. Write the twenty answers your customers genuinely ask for before launch, then read the transcripts for the first couple of weeks and fix what the AI got wrong. That habit matters more than the model behind it.
The third risk is handoff. Automation that cannot gracefully pass a conversation to a person, with the full thread attached, turns into a trap your customers learn to resent. Test this deliberately during your trial: ask for a human, ask something deliberately outside the knowledge base, and see whether you get a clean handover or a loop.
The fourth risk is a quiet one. Do not rebuild Intercom's complexity in a simpler tool. Teams often port over every tag, rule and automation out of habit, then wonder why the new system feels heavy too. Start with a shared inbox, your real channels and a small knowledge base. Add process only when a specific failure demands it.
What is the practical way to run a trial in a week?
Pick two tools, not five. More than two and you will not give any of them a fair test.
On day one, connect only the single channel that carries most of your volume, and point the AI at your website and your existing help content. On day two, write or tidy the answers to your most common questions. For the rest of the week, run real traffic through it and invite every person on your true seat list to watch the inbox, including the ones you would not normally pay for. That last step is the point of the exercise: you are testing the working pattern, not the feature list.
At the end of the week, ask three things. Did customers get correct answers without a person stepping in? When a person had to step in, did they have the context they needed? And did having everyone in one place make the work feel lighter or noisier? Those answers will separate the two candidates faster than any comparison table, and they are the only evidence that reflects your business rather than someone else's.