Skip to main content
Status: Decided — July 3, 2026 Pre-launch context: No live users yet. App is on staging. 0revenue.Purchaseshavenotbeenprovenend−to−end.∗∗Modelofrecord:∗∗[PricingStrategy](/planning/pricing/strategy).Tiersandpricesherefollowthatpageexactly:Guest/Free/∗∗Pro0 revenue. Purchases have not been proven end-to-end. **Model of record:** [Pricing Strategy](/planning/pricing/strategy). Tiers and prices here follow that page exactly: Guest / Free / **Pro 12.99/mo, $99.99/yr, 7-day trial**; Studio deferred to v4.0.

​
The shape of the rollout

What this plan explicitly drops from earlier drafts: the Day-1 Studio tier, “no free trials, ever,” 9.99Propricing,HistoricalImportasone−timeSKUs,andthe600−unit/9.99 Pro pricing, Historical Import as one-time SKUs, and the 600-unit / 149K Founders program. Those live on only as rejected alternatives in Pricing Strategy. The smaller 50-unit Founders lifetime experiment is likewise cancelled — no lifetime products of any kind (founder decision, July 3, 2026). Also corrected: Apple refunds are discretionary, case-by-case decisions by Apple — there is no guaranteed “14-day refund window,” so refunds can never be our only risk-reversal story. The 7-day trial is the risk-reversal mechanism.

​
Phase 0 — Prove the purchase machine (current, BLOCKED)

Nothing else in this document matters until a stranger can pay us and we can see it happen. Phase 0 is engineering + configuration work, currently blocked — track and clear these before any launch date talk:
1

RevenueCat dashboard configuration

Products, entitlement (pro), offering, and 7-day introductory trial configured in RevenueCat + App Store Connect, matching the SKU block at the bottom of this page. Verify the trial constant in code matches the store-configured intro offer.
2

Sandbox purchase, restore, and webhook verification

Full loop in the App Store sandbox: purchase (monthly + annual), trial start → conversion, restore purchases on a fresh install, cancellation, and RevenueCat webhooks landing and updating entitlement state server-side. Include the billing-grace path: a 16-day billing grace period already exists in code — exercise it in sandbox.
3

TestFlight paid beta

A small external TestFlight cohort runs the real paywall (sandbox billing) end-to-end: onboarding → Free account → trial start → first AI Notebook inside the trial week → conversion or lapse. This is also the first real read on the trial experience.
4

Apple Small Business Program enrollment

Enroll immediately so Apple’s cut is 15% instead of 30% from day one. Until confirmation, all models use 30% Year-1 economics (see Pricing Strategy).
Exit criteria: every step above green, plus refund/cancellation states rendering correctly in-app.

​
Phase 1 — App Store launch

Guest + Free + Pro. One paid tier, one entitlement, a 7-day trial.

​
Pricing

Historical Import is included in Pro with usage caps (daily order-processing rate cap + per-account AI-generation ceiling). No import SKUs at launch.
Why a trial: High IQ is an unknown brand asking $12.99/mo. The trial lets the first AI Notebook sell the subscription, and it is the honest risk-reversal offer — unlike Apple refunds, which are discretionary and cannot be promised. Feature previews on Guest/Free (strain encyclopedia, label scanner, music) remain the pre-trial taste of the product.

​
What Phase 1 ships

The full launch feature-by-tier matrix lives in Tier Comparison — canonical, not duplicated here. In one line: Guest browses, Free tracks (stash / favorites / manual orders / share cards), Pro thinks (Notebooks, Stats, Gmail sync, receipt parsing, Suggest Strain, capped Historical Import).

​
What Phase 1 specifically does NOT have

  • No Studio tier and no AI media (v4.0 — Phase 3)
  • No Basic tier, no à-la-carte add-ons, no family or B2B plans
  • No Historical Import SKUs (included in Pro, capped)
  • No lifetime products, period — founder decision (July 3, 2026); early-adopter marketing uses non-pricing perks (founder badge, early access, direct line to the founder)

​
Launch metrics to watch

Targets and thresholds are defined in Revenue Projections — dashboard: install→trial (≥10%), trial→paid (≥40%), net install→paid (3–5%), cohort churn (≤7%), annual mix (25–35%), refund leak (2–5%), AI cost/user (≤$2.50/mo).

​
Phase 2 — Optimization experiments (~90 days post-launch)

Once real cohort data exists, run experiments in priority order. Each is a test, not a decree.
1

Paywall A/B testing

Placement (onboarding vs feature-gate), copy, price framing, and trial presentation. RevenueCat Experiments handles the split. Highest-leverage lever per the sensitivity analysis in Revenue Projections.
2

The no-trial test (rejected Option B, revisited with data)

Trigger: trial→paid below ~30% for two consecutive cohorts, or material trial abuse. Test the rejected “Guest + Pro only, no trial” model against the control. This is the sanctioned way Option B re-enters — as an experiment, not a relaunch. See Pricing Strategy — rejected alternatives.
3

Annual push

Default the paywall to annual, test discount framing (“$8.33/mo billed yearly”), and post-conversion monthly→annual upgrade prompts. Target: 25–35% annual mix.
4

Win-back and dunning

Current state: a 16-day billing grace period exists in code, but there is no dunning flow yet. Build: payment-failure messaging during grace, lapsed-subscriber win-back offers (RevenueCat win-back offers on iOS), and a cancel-flow save offer. Irrelevant at 40 subscribers, compounding at 1,000 — build it during Phase 2, not after churn hurts.
Also in Phase 2, decided by data rather than scheduled: whether the churn feedback justifies a cheap Basic tier (rejected at launch), and whether Historical Import caps need tuning or a deep-archive add-on experiment.

​
Phase 3 — Studio (v4.0, 12+ months post-launch)

Trigger: AI media features (podcasts, video reports, strain art, session music) are production-ready and the Pro funnel is proven (Stage 2 of Revenue Projections reached or in sight). Decisions deliberately deferred to Phase 3 planning, with real data:
  • Tier vs add-ons vs hybrid for AI media (the analysis preserved in earlier drafts of Pricing Strategy gets re-run with actual generation costs)
  • Studio trial policy, upgrade-hook design (15-second previews, blurred/watermarked media below reports — that design survives from the brainstorms)
  • Whether the AI Copilot concept anchors Studio — noting its economics only work usage-capped or at a ≥$24.99 price point
  • Pro→Studio upgrade targets (the old flat “15% upgrade” assumption is retired; set targets from observed engagement with previews)
Why Studio is not at launch: selling a $19.99 tier on the promise of future AI media, before a single proven Pro conversion, invites refund requests and review damage and splits an unproven paywall. Studio ships when its features ship.

​
TL;DR — what to build in RevenueCat (canonical SKU plan)

Phase 1 launch SKUs (2 subscription products, 1 entitlement — no one-time products; lifetime SKUs are cancelled, founder decision July 3, 2026):
Phase 2 (no new SKUs): RevenueCat Experiments for paywall A/B and the no-trial test; win-back offers configured in App Store Connect; dunning flow built app-side on the existing 16-day grace period. Phase 3 SKUs (Studio, v4.0):