QUALITY ASSURANCE
BY DAVID KARANDISH, CAPACITY
ILLUSTRATION PROVIDED BY ADOBE STOCK
WHY QA HAS TO CHANGE
AI IS FORCING NEW AUTOMATED PROCESSES.
Contact centers are on the front lines of AI adoption. Salesforce research found that AI agent implementation in customer service organizations jumped from 39 % to 66 % between 2025 and 2026. Also, 77 % of service teams with AI agents now deploy them across both customer-facing and internal operations.
In some contact centers, there are more AI agents than humans running the phone lines and answering emails. Yet many organizations still measure quality as if humans were doing all the work.
Contact centers already struggle to use the data they collect: 96 % record calls but 73 % don ' t have time to analyze them. With agents deflecting and handling more volume, the traditional quality assurance( QA) methods are becoming even less practical.
QA can no longer exist primarily to grade human representatives after interactions end.
QA IS NOT JUST FOR GRADING HUMAN AGENTS
Maintaining quality in customer service has been, and in many organizations still is, an incredibly manual process.
• A manager or QA analyst pulls a sample from a recorded call, works through a scorecard, drafts a brief evaluation of the representative, and files the feedback away for the next coaching session or 1:1.
• Those insights support the manager’ s relationship with the team member, guiding their career pathing and professional development goals.
• Those improvements, ideally, trickle down to customer experience( CX).
Not only is a manual process like this incredibly time-consuming from an operational standpoint, but it also only makes sense when the human representative is the one driving the customer interaction.
Today, a customer is more likely to interact with some form of automation or virtual agent before they ever reach human support. They may even get passed from AI to human, back to AI, before reaching a true resolution.
NOVEMBER 2026 37