KPI benchmark

What is a good containment rate for an AI agent?

Across 56 public AI deployments, the median containment rate is 70%, and the middle half of them report between 49.3% and 85%. Share of conversations the AI resolves without a human taking over.

70%

Median containment rate

49.3% to 85%

Middle half of deployments

56

Public deployments pooled

23 of the 56 values are claims by the organization itself, a regulator or an independent study; 33 come from the AI vendor. One value per organization; "up to" claims are listed but not pooled. Data as of 11 November 2026. Methodology

How to measure it

Resolved conversations without handover divided by all conversations the AI handled, over the same period. Count a conversation as contained only if the customer did not come back on the same topic within seven days.

Across the library

Median 70% across 56 deployments, reported range 20% to 99.6%, plus 1 "up to" value not pooled.

Higher is better. Unit: percent.

Reported values by use case

AI agent for first line contact center service

AI agent for account and card servicing

AI agent for IT service desk resolution

AI agent for parcel tracking and delivery exceptions

AI assistant for HR and policy questions

AI voice agent for drive through and phone order taking

AI agent for insurance policy servicing

  • Lemonadeat least 50% · organization claim
  • LAQO30% · vendor claim

AI agent for order status, delivery changes and returns

  • Nuuly35% · vendor claim
  • NOBULLabout 30% · vendor claim

AI travel and hotel booking concierge

Use cases that should track containment rate

Frequently asked questions

What is a good containment rate for an AI agent?
Across 56 public AI deployments, the median containment rate is 70%, and the middle half of them report between 49.3% and 85%. Share of conversations the AI resolves without a human taking over.
How do you measure containment rate?
Resolved conversations without handover divided by all conversations the AI handled, over the same period. Count a conversation as contained only if the customer did not come back on the same topic within seven days.
How does containment rate differ by industry?
The median containment rate is 75% in banking (11 deployments), 83% in travel and hospitality (8 deployments), 70% in telecommunications (7 deployments), 70% in energy and utilities (5 deployments), 70% in media and entertainment (5 deployments), 72.5% in logistics and transportation (4 deployments), 38.5% in retail and ecommerce (4 deployments) and 65% in insurance (3 deployments). Industries with fewer than 3 reported values are not shown.
Which AI use cases report the best containment rate?
Among use cases with at least 3 reported values: AI assistant for HR and policy questions (median 94%, 3 deployments); AI agent for parcel tracking and delivery exceptions (median 85%, 3 deployments); AI agent for account and card servicing (median 84.8%, 4 deployments); AI agent for device and connectivity troubleshooting on voice and chat (median 70%, 3 deployments).
Do AI vendors report a different containment rate than the organizations themselves?
Claims by the organizations themselves, regulators or independent studies have a median of 73% (23 values); claims by the AI vendor have a median of 67% (33 values). The page lists both, and every value links to its source.
Where do these containment rate numbers come from?
From public AI deployments in the Blits.ai AI Use Case Library. Every value is quoted from a named source, each organization counts once, and "up to" claims are listed but left out of the median. Data as of 11 November 2026.