Most teams treat these six screens as visual polish — something to tidy before launch. The research says the opposite: each one has at least one published result where a copy or structure change moved real money, at a size most roadmap bets never reach. This is the case for owning them deliberately, one screen at a time.
Telemetry
Money screen 01 · The paywall
Buyers defuse your trial the day it starts.
In vendor subscription telemetry across 115,000+ apps, 55 of every 100 three-day-trial cancellations happen on day zero — and for 7-day trials it is still roughly 40 of every 100. People start the trial and immediately kill the billing, not because the product failed, but to protect themselves from a forgotten charge. The paywall's real job is answering that fear in writing: a visible charge timeline — today, the reminder day, the charge day. Blinkist rebuilt its paywall around exactly this transparency and reported 23% more trial starts — a company-reported figure, labeled as such.
RevenueCat, State of Subscription Apps (vendor telemetry, limits stated) · Blinkist paywall case study (company-reported)
Solid
Money screen 02 · The pricing table
The middle option wins share — in 22 experiments.
The compromise effect is one of the most replicated results in pricing research: across 22 pooled experiments, making an option the middle of the lineup lifted its share — and unlike its famous cousin the decoy effect, it survives realistic product presentations instead of falling apart outside plain number tables. The field has caught on: in our audit of 60 top SaaS pricing pages, 83% of the pages that highlight a plan highlight an interior tier — neither the cheapest nor the most expensive.
Neumann et al., pooled compromise-effect experiments · Science-Based Software audit of 60 SaaS pricing pages (our audit, graded as such)
Solid
Money screen 03 · The first run
Value before identity — and a reduced interface to get there.
Duolingo's most famous onboarding win was deleting a screen: moving account creation from the front door to after the first lesson raised daily active users by about 20% in its published A/B tests, and softening the wall further added another 8.2% — company-reported, direction consistent with the commitment literature. Underneath it sits one of HCI's oldest experiments: learners given a reduced "training wheels" interface finished the standard tasks in 78 minutes against 107 — 27% faster, with better comprehension that persisted after the full interface unlocked.
Gotthilf / First Round Review (company-published A/Bs) · Carroll & Carrithers (1984), Communications of the ACM — controlled experiment
Solid
Money screen 04 · The upgrade prompt
The biggest randomized test in trials says: ask sooner.
In the largest trial-length experiment ever published — 337,724 users randomly assigned to 7, 14, or 30-day trials at a real SaaS firm — the 7-day trial won on every money metric: +5.6% subscriptions, +7.9% revenue, and better retention two years out. Urgency beat exploration time. The upgrade prompt is not an interruption to delay as long as politely possible; the randomized evidence says the shorter window converts more people and loses nothing downstream, provided a new user can reach first value inside it.
Yoganarasimhan et al., randomized field experiment, n=337,724 — peer-reviewed
Solid
Money screen 05 · The signup form
Asking is an intervention, not a container.
In a study of over 40,000 households, families asked a single purchase-intent question — no ad, no pitch — went on to buy cars at 3.3% versus 2.4% for those never asked. The mere-measurement effect has since been meta-analyzed across 116 tests: small, consistently real, and strongest when people predict their own behavior. Every field on your signup form is a treatment applied to the user, not a box that passively collects data — which is why questions that visibly change what happens next convert, and questions that feed a marketing profile quietly cost you signups.
Morwitz, Johnson & Schmittlein (1993), Journal of Consumer Research · Wood et al. (2016) meta-analysis, 116 tests
Regulatory
Money screen 06 · The cancel flow
The easy cancel is now the law — and the retention strategy.
California's AB 2863 took effect July 2025: cancellation must work in the same medium as signup, with renewal reminders and recorded consent. The federal click-to-cancel rule died on procedure days before its deadline, but enforcement under ROSCA never paused — so California's standard is the de facto US baseline. The commercial logic points the same way as the legal one: the day-zero self-protection in screen 01 is what buyers do when they expect cancellation to be hard. A two-tap cancel is what makes "cancel anytime" true — and what makes people willing to start the trial at all.
California AB 2863 (effective July 1, 2025) · Eighth Circuit vacatur of the FTC rule + continuing ROSCA enforcement