Deflection Ceiling

The Deflection Ceiling is the share of inbound volume a triage layer can absorb before accuracy, tone, or escalation quality degrades. It is not a vendor setting; it is a property of the client's ticket mix. Agencies that price against a promised deflection rate rather than a measured ceiling end up staffing the gap themselves, which converts a cost-reduction retainer into a hidden labor subsidy. The practical move is to instrument the ceiling before quoting: run a two-week baseline on the top five intent clusters, then commit only to the volume those clusters represent. Forethought's Triage Agent and Julia 1's CPU-based routing across 52 locales both classify and rank requests, but neither removes the long tail of emotionally charged or multi-system cases. One agency selling a 60 percent deflection promise on a mix where only 38 percent of tickets are single-intent will absorb the difference in unbilled hours, and that gap is where the margin goes.