Shrinking call-centre employment is a mobility problem, not only a headcount signal

Revelio Labs reported on 6 October that global call-centre headcount was 5.3% below its December 2023 peak after eight consecutive quarters of year-on-year decline. Its analysis of professional profiles and firm headcounts says only 10.8% of movers reached technical support, customer success, software or data roles, where median pay was higher. The Financial Times independently covered the analysis, noting that the decline appears to reflect weaker hiring more than a wave of layoffs.

Measure the bridge, not just the fall

The timing coincides with the spread of AI chatbots, but coincidence is not causal identification. Offshoring, demand, firm mix and classification changes could also affect headcount. Professional-profile data can miss informal workers and lag job changes. The numbers therefore support a labour-market signal, not a clean estimate of jobs lost to AI.

The stronger decision point is mobility. Before automating a customer-service queue, map the destination roles that actually exist in the same labour market. Define the skills gap for technical support, customer success, quality operations and data stewardship; then reserve paid practice time, supervised work and hiring slots. A course completion badge is not a transition if no receiving team accepts the evidence.

Track three cohorts for at least a year: workers whose tasks changed, workers who moved internally and workers who exited. Report destination role, pay band, retention and time to competent performance. If the programme raises training participation but movers still land in lower-paid service work, the bridge is not working. Headcount efficiency should not be counted without the cost and distribution of the transition it creates.