Switching From Intercom Without Breaking Support

Why people start looking in the first place

Intercom runs two meters at the same time. The first is per-seat pricing. Per-seat pricing means you pay for every person on the team, whether they answer one chat or a hundred. Intercom Essential is $39 per seat per month on monthly billing, so ten seats is $390 before anyone has used AI at all. We checked that figure in July 2026, so confirm it against Intercom's current published pricing before you build a business case on it.

The second meter is Fin, Intercom's AI agent. Fin is billed at $0.99 per resolution. A resolution means a conversation the AI finished on its own, with no human stepping in. So a month where the AI resolves 500 conversations adds roughly $495 on top of whatever the seats already cost.

Put the two together and the complaint writes itself. The two things you most want, more people on the rota and more automation, both push the invoice up. A good month for the AI is an expensive month for you, and a bill you cannot forecast is what usually starts the search.

There is a quieter second reason. In June 2026, Intercom announced on its own blog that Salesforce had signed an agreement to acquire Fin, its AI agent business. An acquisition is not automatically bad news. But it does add roadmap uncertainty around the exact feature many teams are paying the most for, and that is a fair thing to weigh at renewal time.

Step 1: export everything, before you change any setting

Do this while your account is still fully active. Downgrading or cancelling first can restrict access to data you will want later.

Pull four things. Conversation history, through CSV export and the API. Help centre articles, which matter most, because your articles are what any new AI learns from. Contacts and companies, as a plain CSV. And saved replies, which have no clean export, so copy them into a document by hand.

Then verify. Open the files and check the row counts against what the dashboard says you have. A truncated CSV you discover three months later is not recoverable, and this five-minute check is the cheapest insurance in the whole project.

Step 2: know exactly what breaks

The rule of thumb is simple. Data moves. Logic does not.

Rebuilding automations sounds worse than it is. Most teams configured far more rules than they use, so a rebuild is a free chance to bin the ones that fire on nothing. Start with your top twenty real intents and stop there.

Step 3: run both tools side by side

This is the part that makes a switch safe, and it is the part teams skip. One rule first: never load two chat widgets on the same page. Two launchers in the corner confuse visitors and ruin your comparison data at the same time.

There are two clean ways to split. Split by page, putting the new widget on your docs, blog and marketing pages while Intercom keeps the logged-in app. Or split by traffic, sending 10 to 20 percent of visitors to the new widget through your tag manager. Page splitting is simpler. Traffic splitting is fairer, because both tools see the same mix of questions.

One to two weeks is enough. Intercom still carries the main load, so nothing customer-facing is at risk while you find the gaps in your new setup.

Step 4: test the five things that actually matter

Not the feature list. Feature lists all look the same. Test these instead.

1. Accuracy on your real questions. Take the 50 most common questions out of your conversation export and run them through the new AI. Not sample questions. Yours.
2. The handoff. What happens when the AI does not know? It should say so and pass the conversation to a person with the full history attached. A confident wrong answer is worse than no answer.
3. Where your team replies. EzyConn conversations arrive in Slack and Microsoft Teams, and agents answer from the Slack thread rather than a second dashboard. Check whether your team actually prefers that before you assume it.
4. Mobile. Most chat traffic is on a phone. Open the widget on a real handset, not just a resized browser window.
5. The bill in a bad month. Model your worst month, not your average one. Double the AI volume and see what each tool charges.

How long the whole thing takes

When to switch from Intercom

If an annual deal has six months to run, switching means paying for two tools at once. Time the move to the renewal date instead.

Until then, run the side-by-side on part of your site. EzyConn answers from your own website pages and documents, shows the source, and hands the chat to your team when a visitor asks for a person, so the comparison is on your real traffic.

What the other side looks like

If you do move to EzyConn, the pricing shape is the thing that changes. There are no per-seat fees and no per-resolution fees. AI runs on GPT-4o and Claude and is included on every plan, including the free one, so the AI answering more questions never costs you more.

The free plan is free forever, not a trial: 2 seats, 500 messages a month and one chatbot. Starter at $25 a month carries 3 agent seats plus the owner, and up to 10 agent seats come with Professional at $95 a month, which is the tier most switching teams land on. Annual billing takes 20 percent off. Full detail is on the pricing page, and the line-by-line comparison lives on the Intercom alternative page.

Questions people ask before switching

Intercom pricing figures quoted here were checked in July 2026 and describe monthly billing. Vendor pricing changes, so verify current published rates before deciding.