CAST AI
CAST AI optimises customer cloud spend by automatically reducing waste in Kubernetes clusters. The product only works if it generates real, measurable savings. This characteristic guided an unusual KPI choice.
Instead of measuring internally by number of customers or ARR, CAST AI publishes and monitors a metric it calls "cloud savings delivered" — the aggregate sum of savings generated for all active customers. A number that only grows if the product actually works.
This KPI has three effects. First: it aligns the product team on the metric that matters to the customer (actual savings), not on proxy metrics that might optimise vanity without real value. Second: it becomes a powerful marketing argument — "we have saved our customers $X in cloud spend" is more credible than any performance benchmark. Third: it naturally guides pricing: CAST AI has adopted a model where customers pay a percentage of savings generated — perfectly aligned incentives.
The lesson: the most effective KPI is the one that measures the value the customer receives, not the value the company captures. When these two numbers are correlated, growth becomes almost automatic.
A startup's ideal KPI is the one that increases only when the customer succeeds. Finding it means finding perfect alignment between company incentives and customer value.