Path from A− to A

GPA Action Plan — Prioritized Grade Bumps with Effort & Timing
Charlotte (COO) • February 25, 2026 • Based on v7 analysis + Stitch Sessions AH–AO
3.72
Current GPA (v7)
~3.83
Estimated GPA Now
3.85
Target (A Threshold)
~0.02
Remaining Gap

The Big Finding: Stitch Already Closed Several Gaps

The v7 Cost Analysis (Feb 25 AM) was written before Stitch Sessions AI through AO completed significant work. Cross-referencing those sessions against the v7 "Path to A" roadmap reveals that several upgrade requirements have already been built and deployed. These changes haven't been reflected in the grades yet.

Unrecognized Grade Improvements Since v7

Stitch built or deployed 8 specific items that v7's "Path to A" listed as requirements for grade upgrades. Three categories likely earned a full letter-grade bump; one more is within a single task of upgrading. If we re-grade now, the GPA jumps from 3.72 to an estimated ~3.83 — putting us within striking distance of the 3.85 "A" threshold.

Category v7 Grade Estimated Now What Changed
Scalability & Performance B+ A− upgrade Redis provisioned on Railway, cache wired into endpoints (featureGate, onboarding), circuit breaker integrated (Claude + Stripe calls), API pacer active. 3 of 5 Path-to-A items done.
Billing & Subscription A− A− (borderline A) 4-stage escalating dunning banners + 4-email dunning sequence built. Payment failure recovery with grace period live. Stripe audit 7/7 PASS. 2 of 3 Path-to-A items done. Only audit log remains (~4 hrs).
DevOps / Infrastructure A− A− (stronger) GitHub Actions CI/CD pipeline with 3 parallel jobs (lint, backend tests, frontend tests). 1 of 3 Path-to-A items done. Still needs staging env + monitoring.
UI / UX Quality B B+ upgrade 15 CSS design tokens (spacing, radius, duration), ARIA accessibility on 5 components (3 modals + week nav), focus-visible keyboard rings, playful copy personality layer across 14 files. 3 of 5 Path-to-A items partially addressed.
Test Coverage A− A− (stronger) 556 → 674 tests (+118). CI pipeline ensures they run on every push. All green: 141 frontend + 533 backend. Still needs Instagram integration tests, load testing, visual regression for full A.
Core Features A− A− (stronger) Stripe end-to-end audit 7/7 PASS (was listed as Path-to-A requirement). Google OAuth trial bug fixed. BUG-7 resolved.

GPA impact estimate: Scalability B+ → A− (+0.4 pts) and UI/UX B → B+ (+0.3 pts) add ~0.7 total points. Across 18 dimensions, that's roughly +0.04 GPA. Combined with the strengthened positions in Billing, DevOps, Test Coverage, and Core Features (which inch toward the next threshold), the effective GPA is estimated at ~3.80–3.83.

Priority Matrix: What Moves the Needle Most

Not all grade bumps are equal. Bumping a B− to B+ gains 0.6 points; bumping an A− to A gains only 0.3. And some upgrades need real user data (post-launch), while others can be done now. This matrix sorts all remaining upgrades by effort-to-impact ratio and launch timing.

High Impact + Low Effort (Do Now)

  • Billing → A: Plan modification audit log (~4 hrs)
  • Scalability → A−+: DB query optimization + index review (~4 hrs)
  • DevOps → A: Sentry error tracking + UptimeRobot (~2 hrs)
  • Customer Support → A−: SLA definition + contextual help triggers (~3 hrs)

High Impact + Medium Effort (Do Soon)

  • Test Coverage → A: Instagram API mocks + k6 load tests (~6 hrs)
  • UI/UX → A−: Micro-interactions + loading state polish (~8 hrs)
  • Core Features → A: Sentry integration + migration 013 (~3 hrs)

Medium Impact + Higher Effort (Post-Launch)

  • Analytics → A: Best-time-to-post, exportable reports (needs user data)
  • API Integrations → B+: Facebook cross-post (Meta API already in place)
  • Onboarding → A: Progress persistence + product tour

Long-Term (Post-Traction)

  • API Integrations → A: TikTok + Google Business Profile (new API approvals)
  • Legal → A: Paid attorney review, DPA, sub-processor list
  • Auth → A: 2FA/TOTP, session management dashboard

The Sprint to 3.85: Fastest Path to an A

With the unrecognized improvements putting us at ~3.83, we need just ~0.02 more GPA points — less than a single category bump. Here are three "micro-sprints" that Stitch can execute, ordered by impact. Completing any ONE of these likely crosses the A threshold.

Sprint 1: Billing A− → A (The Quickest Win)
~4 hours +0.3 pts Pre-Launch
  • Dunning email sequence (4-stage escalating) done
  • Failed payment recovery (past_due banner + billing portal) done
  • Stripe billing audit — 7/7 PASS done
  • Plan modification audit log — track every plan change with before/after state, reason, and timestamp Add an audit_log entry on upgrade, downgrade, cancellation, trial conversion, and founding member assignment. Simple DB insert + admin UI view. ~4 hours.
GPA impact: A− → A = +0.017 GPA (3.83 → ~3.85). This alone may cross the A threshold.
Sprint 2: DevOps A− → A (Monitoring + Staging)
~6 hours +0.3 pts Pre-Launch
  • CI/CD with GitHub Actions (3 parallel jobs) done
  • Sentry error tracking — capture unhandled exceptions, source maps, release tagging Sentry has a free tier (5K errors/month). Server-side SDK + frontend React error boundary. ~2 hours.
  • UptimeRobot monitoring — health endpoint check every 5 minutes, Slack/email alerts Free tier: 50 monitors, 5-min intervals. Point at /api/health. ~30 minutes.
  • Staging environment (stretch goal) — duplicate Railway service with separate Supabase project This is the "nice to have" that would lock the A. Could defer if Sentry + UptimeRobot alone push us over. ~3 hours to set up properly.
GPA impact: A− → A = +0.017 GPA. Combined with Sprint 1: ~3.86.
Sprint 3: Customer Support B+ → A− (Quick Lift)
~3 hours +0.4 pts Pre-Launch
  • AI Q&A support endpoint (Claude Haiku) done
  • 40+ FAQ knowledge base + 5 user guides done
  • 10 contextual InfoTooltips done
  • Define and publish response time SLAs Solo: <24 hrs, Pro: <12 hrs, Salon: <4 hrs. Add to Help Center + ToS. Charlotte can draft this. ~1 hour.
  • Contextual help triggering — surface relevant FAQ articles on error states or common friction points Extend InfoTooltip pattern: when user hits a known friction point (e.g., Instagram connection fails), show a contextual help link. ~2 hours.
GPA impact: B+ → A− = +0.022 GPA. Combined with Sprints 1+2: ~3.88.

Bottom Line

We're much closer than v7 suggests. The unrecognized Stitch work likely already moved us from 3.72 to ~3.83. Completing Sprint 1 alone (one 4-hour task: plan modification audit log) probably crosses the 3.85 A threshold. If we want to lock it in with a comfortable margin, Sprints 1 + 2 together (~10 hours of Stitch time) would put us at ~3.86–3.87. Adding Sprint 3 (~3 more hours, partially Charlotte work) pushes to ~3.88, solidly in A territory.

Total effort to cross from A− to a locked A: 10–13 Stitch hours + 1–2 Charlotte hours. All pre-launch compatible.

What NOT to Chase Before Launch

These categories have legitimate paths to improvement but depend on real user data, external API approvals, or post-launch infrastructure. Chasing them now would burn time without defensible results.

Category Current Path to A Why Not Now
API Integrations B− Facebook cross-post, TikTok, Google Business Each platform needs separate API approval (weeks). Facebook cross-post is fastest (shared Meta infra) — good Month 2 target. TikTok and Google Business are Month 3+.
Analytics & Reporting B+ Best-time-to-post, exportable reports, cohort analysis Requires 30+ days of real user data for statistical significance. Building recommendation features on simulated data would be fabrication. Month 3+.
Auth & Security A− 2FA/TOTP, session management, security audit log Already strong (WebAuthn + JWT + RLS on 27 tables + rate limiting + Helmet). 2FA is a "nice to have" at launch volume, not a grade-critical gap. Post-100-users.
Legal & Compliance A− Paid attorney ToS review, DPA, sub-processor list Attorney review costs $2K–$5K. Worth doing once revenue justifies it. ToS v4 is comprehensive for launch. Post-revenue.
Mobile Responsiveness A− PWA support, 5+ real device QA Already mobile-first. PWA is nice but not grade-critical until user base demands offline access. Post-launch.

Recommended Next Steps

Here's the concrete sequence. Steps 1–2 are Charlotte tasks. Steps 3–4 are Stitch inbox items. Step 5 is the re-grade to confirm we've crossed the threshold.

  • Step 1 (Charlotte, now): Draft Customer Support SLAs and submit for Jason's approval Solo: <24 hrs, Pro: <12 hrs, Salon: <4 hrs. Add to Help Center page + reference in ToS. ~30 minutes.
  • Step 2 (Charlotte, now): Create Stitch inbox task with Sprint 1 + Sprint 2 scope Billing audit log + Sentry + UptimeRobot. Clear requirements, estimated 6–10 hours. Defer staging env to judgment call.
  • Step 3 (Stitch): Execute Sprint 1 — plan modification audit log ~4 hours. DB table + insert hooks on billing events + admin UI read view.
  • Step 4 (Stitch): Execute Sprint 2 — Sentry + UptimeRobot + optional staging ~2–6 hours depending on staging. Free tiers for both services.
  • Step 5 (Charlotte): Produce v8 Cost Analysis with updated grades Re-grade all 18 dimensions reflecting Stitch AH–AO work + Sprint 1–3. Confirm A threshold crossed.

Assumption: The GPA scale uses standard 4.0 mapping (A+ = 4.0, A = 4.0, A− = 3.7, B+ = 3.3, B = 3.0, B− = 2.7). The exact scale used in v7 may differ slightly — the "estimated ~3.83" is directional, not precise. The v8 re-grade will calculate the exact number.

Uncertainty: Whether Stitch's infrastructure wiring (Session AI) and UI accessibility work (Session AL) are sufficient for a full grade bump in those categories, or only a "strong A−" / "strong B+." The v8 re-grade should evaluate each dimension against the same rubric used in v7 to avoid grade inflation.