Gartner: Returns are unclear for 42% of AI use cases in customer service
Of the 432 use cases Gartner examined, about a quarter delivered a positive return and about a quarter a negative one; returns are unclear for 42% of use cases, and 11% broke even.
What's happening
According to Gartner's analysis, about a quarter of the 432 use cases examined delivered a positive return and about a quarter a negative return. For 42% of the use cases, the return is unclear: leaders say they do not know what value was generated. Another 11% broke even. Teams run about 5 use cases on average and devote about 13% of their function's budget to AI. Among leaders, 56% expect their bonuses to be tied to AI results in 2026, and more than three-quarters plan to increase AI investment in 2026.
Where it comes from
According to Info-Tech Research Group, the problem is a top-down approach that starts with board pressure rather than customer need.
Whose desk it lands on
Customer service directors and call center operations managers should decide which customer metric they will use to measure the value each of their AI use cases generates.
Our view
The CX Brief's interpretation, not a claim made by the source.
In Gartner's data, use cases with unclear returns make up the largest single group. A customer service leader in Türkiye who asks for an AI budget can be expected to explain, along with the request, how they will demonstrate the value the spending is meant to generate.
Worth noting
The "unclear" category covers use cases for which leaders say they do not know what value was generated; it does not mean those use cases failed.
