Clay vs Planable

Clay ships a weekly roundup rather than individual release notes, and the last two windows point in one direction: proving and controlling spend. Reporting in Audiences now shows pipeline and closed-won revenue from Clay-sourced accounts, credit usage gets time-series graphs and near-real-time dashboards, and CPJ allowance consumption is tracked in-product. Around that sit steady platform widenings — filterable enrichment-provider search, a Connections page that traces where each connection is used, higher field limits in Audiences, and CLI device login for headless environments.

Two arcs are converging. Claygent is being pushed toward cheaper inference with open-weight models and a distinct product identity, while the surrounding platform is being instrumented so buyers can see exactly what credits bought them. That combination reads as a response to the central objection facing consumption-priced GTM tooling: teams can spend fast and struggle to attribute the outcome. Clay is also spreading outward into other agents' surfaces rather than keeping work inside its own tables.

Expect attribution reporting to deepen — most likely per-workflow or per-function revenue attribution rather than only account-level — and expect more model choice inside Claygent as open-weight options prove out on cost. The roundups suggest Audiences is where the near-term investment is going.