SeedSweep

Containment rate vs resolution rate

Containment rate is the share of conversations that ended without reaching a person. Resolution rate is meant to be the share where the customer's problem was solved. A conversation can be contained without being resolved, for example when the customer gives up and calls instead.

Last checked 30 September 2026 by SeedSweep.

What each rate counts

Containment rate is the number of conversations that never went to a person, divided by all conversations the bot handled. It tells you how much work the bot kept away from your team. It tells you nothing about whether the answers were right.

Resolution rate is the number of conversations where the problem was solved, divided by all conversations. The hard part is the word "solved". Nobody asks every customer, so each vendor writes its own rule for what counts, and that rule decides the number.

You will also see automation rate. It usually means the share of work the AI finished without a person. Fin's glossary defines it as involvement rate times resolution rate, so it inherits whatever the resolution rule counts.

How a vendor decides a conversation was resolved

Intercom's Fin counts two kinds of resolution. A confirmed resolution is one where the customer says the answer helped. An assumed resolution is one where, after Fin's last answer, the customer "exits the conversation without requesting further assistance". Fin bills per outcome, and Intercom's help center says an assumed resolution is charged, so the same rule also sets the bill.

Assuming a resolution is a fair guess when someone asked how to reset a password and left. It is a poor guess when the bot quoted last year's refund policy and the customer left to dispute the charge with their bank. On the dashboard both look the same.

Why the gap matters

If the resolution rate in front of you comes from the vendor's dashboard, it follows the vendor's rule. When assumed resolutions make up a large share, the number tells you how often customers stopped talking to the bot, not how often they got a right answer.

The same question comes up when buying a company. In January 2025 the SEC charged Presto Automation over its AI claims. Cooley's summary of the charges says Presto's automation rates counted drive-thru orders completed without restaurant staff involvement "but not without any human involvement". Those are two different counts. If you pay a price based on one count, check that it is the count you meant.

How to check the number yourself

You can run this check with your own team if the person grading did not choose the vendor.

  1. Write down the definitions in use. Find the vendor's rule for resolved, assumed resolved and escalated in its documentation or your contract.
  2. Pull conversations from the last 90 days. Draw them across topics and across outcomes, including ones marked resolved, ones marked assumed resolved and ones that went to a person.
  3. Grade each answer against the help center article or policy that was live when the conversation happened, not today's version. Mark it correct, wrong, based on an outdated article or impossible to tell.
  4. Check what happened next. Look for a repeat contact by any channel within a week, a refund request or a chargeback.
  5. Have a second person grade part of the sample without seeing the first grades, and compare. If the two disagree often, fix the grading rules before you trust any rate.

You end with three numbers side by side. They are the containment rate, the vendor's resolution rate and the share of conversations where the answer was correct and the customer did not come back with the same problem.

An example

Our sample report shows this gap on a support chatbot before a contract renewal. The vendor reported 78% of conversations resolved. In a sample of 400 conversations, 61% of customers had their problem solved, and most failures came from one outdated pricing page. Of the failures, 28% were conversations marked resolved when the customer stopped replying.

The report comes from a real evaluation for a small company, with the details and numbers changed to keep the company anonymous.

What to ask your vendor

  • How is a resolution counted, and are assumed resolutions included in the rate you report?
  • Are assumed resolutions billed?
  • Can we export the conversations behind the rate, with the label each one received?
  • Does a conversation still count as handled by AI if a person approved or edited the answer?
  • Which version of each help center article did the bot use on a given date?

Sources

If a decision rests on this number

If a renewal, a vendor switch or a purchase depends on one of these rates, tell us the decision and its date. We reply within 24 hours.

Contact us