Budgeting app / VP roadmap review
Fix bank access first.
Then earn repeat use.
Focus the roadmap on retaining new users.
38%→45%90-day retention
Company OKR this year
Company OKR this year
Agree to bank access now, and defer broader feature expansion.
Two drop-offs make retention the lens for prioritization
Start with access; repeat use comes next.
Different denominators; these rates cannot be added. Neither identifies a cause or predicts retention lift.
Sequence access, repeat use, then broader retention hypotheses
| Horizon | Retention problem | Scope and confidence |
|---|---|---|
| Now | Reach a usable account | Fix five banks with most failures; manual account entry.Strongest problem evidence; lift unproven. |
| Next | Sustain use beyond week 2 | Weekly summary; categorisation accuracy.Candidates; no causal retention evidence. |
| Later | Explore household and language needs | Shared budgets; Spanish language.Themes to shape, without a build commitment. |
4 engineers · 1 designer · No estimates for any item; horizons express priority, not dates or confirmed capacity.
Access wins now; demand still challenges that order
| Priority | Evidence | Tradeoff or uncertainty |
|---|---|---|
| Bank access now | Five banks: 220 tickets/month. Unsupported banks: 90/month. | Delays shared budgets despite 1,400 app-review votes. |
| Repeat use next | Weekly summary: no data. 12% of transactions miscategorised. | Accuracy problem ≠ evidence for a full engine rewrite. |
| Broader needs later | Shared budgets: top review request. Spanish: 14% of sign-ups’ phones. | Neither establishes an effect on new-user retention. |
Tickets are not unique users; votes are demand signals. Relative effort and retention impact could change the order.
Defer expansion without a demonstrated link to retention
| Not now | Why it yields to retention work |
|---|---|
| Dark mode · CSV export · iOS widgets | 310 / 60 / 180 votes; retention link unproven. |
| Bill reminders · Investment tracking | Partner asks / CEO mention; no quantified demand. |
| Savings goals redesign · Apple Watch app | No data supporting priority. |
| Referral program | Marketing request; no stated retention link. |
The categorisation problem stays next; a full engine rewrite is not a commitment.
Agree bank access now and defer broader feature expansion
- Focus
- Prioritize the five-bank fixes and manual account entry.
- Sequence
- Keep repeat use next; household and language needs later.
- Tradeoff
- Accept delaying shared budgets despite 1,400 votes, plus the deferred list.
- Commitment
- Agree priority order; delivery dates and scope depend on estimates.
The decision serves the 38% → 45% retention OKR; no feature-level lift is established.