When you compare live chat tools, the number that ends up mattering most is not the one printed on the pricing page. It is the pricing model. A tool that looks cheap at $19 per agent can quietly become your most expensive SaaS line the year you grow the team, while a flat plan that looks pricier on day one stays flat as you scale. This is a buyer's guide to telling the two apart.
What per-seat pricing actually costs
Per-seat (or per-agent) pricing charges you for every person you let into the inbox. The headline price is per user, so the real bill is that number times your headcount, every month. The trap is that headcount is exactly the thing that grows when business is going well.
Say a tool is $30 per agent per month. Three agents is $90. A year later you are at ten agents, some of them part-time, and the same tool is $300 a month, or $3,600 a year, for software that costs the vendor almost nothing more to run for ten people than for three.
- →You hesitate before adding the intern who answers three messages a week, because it costs a full seat.
- →You share one login between two people, which breaks every audit trail and assignment feature you paid for.
- →You deactivate seats during quiet months and reactivate them later, turning your support roster into a billing chore.
- →Adding a teammate becomes a finance conversation instead of a one-click invite.
When per-seat makes sense, and when it does not
Per-seat is a fair model when each user gets isolated, individual value: a CRM where every rep has their own pipeline, a design tool where each seat is a creator. There, a seat really does map to value delivered. Live chat is the opposite. It is a shared inbox. The value is the conversation with the visitor, not the number of people who can see it. Charging by viewer count is charging for something that costs the vendor nothing.
How to evaluate a flat-rate plan
If you decide a flat model fits your team, here is the checklist we would run on any vendor, including us:
- 01Are agents truly unlimited on every tier, including the trial? Some plans call themselves flat but still cap seats per tier.
- 02What actually separates the tiers? Good flat pricing segments on things that track the vendor's cost or your scale: number of websites, history retention, SSO, audit. Not headcount.
- 03Can you buy the top plan with a card, or is there a contact-sales wall hiding the real price?
- 04Is AI a forced, per-resolution upsell that re-introduces metered billing through the back door?
- 05Is there a no-card trial and a real refund, so the flat price is reversible if it does not fit?
The two-year test
Do not model the cost at today's team size. Model it at the headcount you expect in twenty-four months, on both pricing models, and compare the totals. Per-seat tools win the first month and lose the second year. If you plan to grow the people answering chats, that second number is the one that lands on your budget.
Pricing is the clearest signal a tool gives you about the relationship it wants. Per-seat says the more of your team uses this, the more you pay. Flat says we built it once, bring everyone. Pick the one that matches how you actually want your team to grow.
“A shared inbox should not cost more because more of your team can see it.”