Where credits actually go
In most workspaces: contact enrichment (emails, phones) dominates, AI columns are growing fast, and firmographics are comparatively cheap. Phone numbers are the most expensive data type; emails are moderate; select validators are free under current pricing.
The levers, in order of impact
- Filter before enrich. Dead rows cost the same as live ones. Suppression and qualification before the first enrichment column is the single biggest lever.
- Right-size waterfalls. Cut steps that add <5% coverage. Order by expected value (match rate ÷ cost).
- Conditional expensive steps. Guard premium enrichments behind fit signals.
- BYO API keys. Where you already subscribe to a provider, BYO keys eliminate that step's credits.
- Stop re-running stable data. Reuse saved enrichment data rather than re-purchasing the same values.
- Watch AI spend. Fixed-price models for routine tasks; token-priced only where synthesis genuinely needs it.
Track the metric that matters
Credits per usable record — total credits ÷ records with valid, usable output. It absorbs match-rate, validity, and waterfall-efficiency effects into one number. Trend it monthly.
Model your scenarios in the Cost & Usage Calculator, including BYO-key savings.
Need help applying this?
Describe your project and the ClayForEnterprises team will review it — the resources stay free either way.