The Contact Center Crossroads: Finding the Right Mix of Humans and AI
Abstract
Based on analysis of millions of customer interactions across 30+ organizations, McKinsey finds that while 50-60% of interactions remain transactional and suitable for automation, the remaining human-handled work will concentrate in complex, judgment-intensive interactions requiring fundamentally different agent capabilities.
★ Curator Summary
McKinsey's data confirms the complexity concentration thesis with real numbers across 30+ organizations. The key finding: AI enables 40-50% fewer agents handling 20-30% more calls. Do the math — that's not a contradiction, it's complexity concentration. Fewer agents, more calls per agent, each call harder. This is the best industry data we have on the operational reality of AI transformation, and it directly supports the three-pool staffing model.
Why It Matters
If 50-60% of interactions are transactional (automatable), your planning needs to account for what happens to the other 40-50%. McKinsey confirms those remaining interactions are 'significantly more complex' — meaning higher AHT, lower FCR, more agent skill required. This is the empirical basis for the complexity concentration parameter in the FOW-Value model: each 10% automation → ~6.5% AHT increase on remaining work.
Caveats
McKinsey industry reports don't disclose detailed methodology. The '30+ organizations' sample is not random — it's McKinsey clients, which skews toward large enterprises. The 40-50% agent reduction projection doesn't account for the second-order demand rebound that will partially offset those savings.
Discussion
Loading...