More tests, better odds
Growth teams live or die on test velocity. Kerf removes the waiting: no engineering ticket for a copy test, no analyst for the readout, no doubt about the numbers.

Where growth teams use Kerf
Landing pages
Test headline, proof and offer on paid traffic with the visual editor, then ship the winner.
Onboarding
Find which step loses people and measure activation, not just completion.
Pricing and paywalls
Test price points, packaging and trial length with revenue as the primary metric.
Lifecycle email
Randomise sends from your ESP and read retention weeks later in the same experiment.
Referral and loops
Measure network effects with cluster level assignment when users influence each other.
Holdouts
Keep a clean group out of every test to measure what growth work adds in total.
Don't ship a winner that costs you later
A discount banner can lift conversion and cut revenue per user. Kerf shows the lift next to guardrails on refunds, support contacts and 90 day retention, so a short term win is labelled as one.
- Revenue and LTV metrics from your warehouse
- Long term holdouts
- Novelty effect detection
