★★★★★
Experiment Signal Lab’s exposure math module finally explained why our Android holdout kept drifting. We rewrote the assignment logic the same week.
Siriporn M. · Experimentation lead
Reviews
What stuck after the cohort ended.
“Short, blunt, useful. The Instrumentation Critique Clinic flagged duplicate purchase events we had ignored for months.”
Pim · Phuket
★★★★☆
Mobile Guardrail Studio is strong on crash pairing. I wanted one more session on offline-first apps; still recommend it for any checkout redesign.
Anonymous client in food delivery
Rated 4.2 on our internal L&D survey after Cohort Relay — mostly because facilitators forced us to kill two underpowered tests before launch.
Platform-style note from a marketplace analytics manager
Case study: paywall copy without a false win
A Bangkok subscription app arrived with three weeks of “+11% start-trial” excitement and a rising refund queue. During Signal Lab reviews we re-centered the primary metric on day-7 retained trial and added refund rate as a hard guardrail.
−38% premature ships blocked
19 days to a clean re-run
+2.1pp retained trials after redesign
The team kept the witty copy but changed eligibility. The win that shipped was smaller — and still online six months later.
Case study: dual-store rollout in Thailand
A lifestyle app staggered Google Play and iOS exposures without documenting the lag. Guardrail Studio’s release calendar exercise caught the mismatch before marketing claimed a country-wide lift.
They left with a shared “store lag” checklist now attached to every experiment brief — low drama, high reuse.