The cost-per-resolution reality check for customer service AI
Gartner now expects generative AI cost per resolution to pass $3 by 2030, and half of companies that cut service staff to rehire. The lesson is not "less AI." It is "aim it better."
Two Gartner predictions from the past month are a useful corrective to the "replace the contact centre" pitch.
First, Gartner predicts that by 2030 the cost per resolution for generative AI in customer service will exceed $3, more than many B2C offshore human agents. In the analyst's words, "full automation will be prohibitively expensive for most organizations."
Second, Gartner expects that by 2027, half of the companies that attributed headcount cuts to AI will rehire for similar work under different titles. Its October 2025 survey of 321 service leaders found only 20% had actually reduced agent staffing because of AI.
Why cost per resolution rises
It sounds backwards, since model prices keep falling. But the easy cases get automated first. What is left is longer, messier, and needs more context, more tool calls and more retries. If you automate a conversation end to end, you pay for every turn of the hard ones too.
Where the economics do work
The back-office calls we automate look different from customer conversations:
- The agent is the caller, not the callee. It phones an insurer to get a status. The script is narrow and the outcome is a few fields.
- Success is binary and checkable. Either the reference number is right or it is not.
- The counterfactual is expensive. The alternative is a trained person on hold.
That is the profile where AI cost per case is a fraction of the manual cost, and stays there.
How we size it
Before we build, we estimate three numbers per workflow: manual cost per case, expected AI cost per case (model, telephony, tool calls, retries), and the share of cases that will still need a person. If the blended cost is not clearly lower, we say so. Sometimes the right answer is an agent that prepares the work for a person instead of one that finishes it.
Automation is not a headcount strategy. It is a cost-per-case strategy, and it should be measured like one.
Sources
- Gartner Predicts GenAI Cost Per Resolution for Customer Service Will Exceed Offshore Human Agent Costs by 2030, Gartner, January 26, 2026.
- Gartner Predicts Half of Companies That Cut Customer Service Staff Due to AI Will Rehire by 2027, Gartner, February 2026.