★★★★★
The Conversion Signal Lab’s module on unequal cohort windows stopped our growth channel report from mixing 14-day trials with weekend promo trials. That alone paid for the tuition in avoided bad launches.
Reviews
Field notes from people who rebuilt their trial stories.
Attribution styles vary because real feedback does. Every quote references work taught inside Portal Scopehub programs.
Useful, though the live critique nights run late if you join from Europe. The paywall moment map still sits on our wall.
We expected magic SQL recipes. Instead we got a ruthless checklist for naming events. Mild letdown at first — then our data warehouse finally stopped inventing “engaged_trial” definitions every sprint.
“Portal Scopehub made finance stop asking for a single conversion percentage and start asking which cohort owed us an explanation.”
Pimchanok R., analytics manager — marketplace ops★★★★☆
Thailand billing appendix clinic was unexpectedly sharp. Wish the Revenue Guild option had more async materials for teammates who missed day two.
Case study: marketplace trial that looked healthy until day nine
A Bangkok marketplace team entered Cohort Forge celebrating a 19% “trial success” rate. After rebuilding cohorts with Portal Scopehub’s equalized window rules, paid conversion among comparable starters sat near 11%. The gap came from promo trials that auto-extended without tagging.
They shipped a corrected dashboard, paused a misleading referral experiment, and redesigned the upgrade prompt that appeared after sellers completed their third listing — a moment identified in the paywall mapping exercise.
Case study: B2B SaaS with two clocks
A security SaaS ran product trials on calendar days while sales-assisted trials used business days. The Trial Lens worksheets forced both clocks onto one timeline. Stakeholders finally agreed that “week one activation” had been two different stories.
Their follow-up experiment tested a mid-trial workspace template rather than another discount. Early qualitative notes suggested clearer intent signals; quantitative readouts remained in progress at the time of writing.