What Should We Put on the Roadmap First in Fintech?
Direct answer: Put first whatever unblocks the customer's most important, most poorly-served job — the moment where users currently work around your product with spreadsheets, screenshots, or a competitor's tab open. In fintech, that usually means the friction point in a money-movement or trust-building job (onboarding, reconciliation, dispute resolution, cash-flow visibility), not the feature your loudest customer requested. Jobs to Be Done (JTBD) forces you to rank by demand for progress, not by feature popularity or internal enthusiasm.
Disclosure: I work on content for Percision (percision.app), an AI strategic-intelligence platform. I'll show where it fits and where a whiteboard and a spreadsheet do the job just as well.
Why fintech roadmaps drift
Fintech roadmaps fail in a specific way. You have three forces pulling at once: compliance and risk (which generate mandatory, non-negotiable work), sales (which promise features to close enterprise deals), and the product team (which wants to reduce technical debt and ship the elegant thing). None of these three natively speak in the customer's language of progress.
The result is a roadmap that's a negotiated truce between departments rather than a bet on where customers most want to move forward. You ship a redesigned dashboard while your users are still exporting to Excel to reconcile transactions — because reconciliation was "phase two" for eighteen months.
Jobs to Be Done fixes this by asking a different question: not "what should we build?" but "what is the customer hiring our product to do, and where are they doing it badly?"
Applying Jobs to Be Done to a fintech roadmap
The core JTBD idea: people don't buy products, they hire them to make progress in a circumstance. A small-business owner doesn't want "accounting software" — she wants to know she can make payroll on Friday without checking three bank apps. Your roadmap should be ranked by how much you improve the jobs your customers are already struggling to complete.
Here's a concrete walkthrough.
Step 1 — Define the jobs, not the personas. Write each core job as: When [situation], I want to [motivation], so I can [expected outcome]. For a B2B payments platform:
- When a large invoice batch comes in, I want to confirm every payment settled, so I can close the books without manual matching.
- When a customer disputes a charge, I want to resolve it fast, so I can avoid chargebacks and keep the relationship.
- When I onboard a new client, I want KYC done in one session, so I don't lose them to a competitor's faster flow.
Step 2 — Score each job on two axes. Importance (how much progress does the customer care about?) and satisfaction (how well is it currently served, by you or anyone?). The gold is high importance, low satisfaction — these are underserved jobs and the strongest roadmap candidates. Well-served jobs are table stakes; leave them alone.
Step 3 — Trace the workarounds. For each underserved job, document what customers do instead. Do they export to Excel? Screenshot and email? Call support? Every workaround is a paid vote that your product isn't finishing the job. In fintech, workarounds also carry hidden compliance and error risk, which strengthens the business case.
Step 4 — Map the job's timeline. Break the job into stages: define, locate, prepare, confirm, execute, monitor, conclude. Fintech products often nail execute (the transaction) but fail at confirm and monitor (did it land? is it reconciled? am I compliant?). The neglected stages are frequently where the biggest roadmap wins hide.
Step 5 — Rank by demand for progress, then sanity-check against constraints. Now overlay reality: regulatory deadlines, security dependencies, and the cost to serve. A high-value job that requires a licence you don't hold moves down. A mandatory compliance item moves up regardless of JTBD score — but you label it as mandatory rather than pretending it's customer-driven.
What "good" looks like: a ranked roadmap where the top items are named as customer jobs, backed by evidence of workarounds, and each carries an explicit outcome metric ("cut reconciliation time," "reduce dispute resolution to under X days"). Anyone reading it can see why item one is item one.
Where Percision fits — and where it doesn't
Once you've done the customer-interview work (JTBD depends on real conversations — no tool replaces that), you have a pile of jobs, workarounds, and rough impact estimates. Turning that into a defensible, board-ready roadmap is where structured analysis helps.
Percision runs your business context through a JTBD-informed reasoning process and produces prioritization logic, scenario analysis (what if you tackle reconciliation before onboarding?), and a board-ready deck with the trade-offs laid out — in minutes rather than a multi-week planning cycle. For a founder or PM prepping a board or investor conversation, that compresses a lot of synthesis. It stays a co-pilot: it structures the reasoning, you make the call.
When you don't need it: if you have three obvious jobs and a clear winner, a whiteboard and an afternoon of customer calls beat any software. If your roadmap is dominated by a single regulatory mandate, that's a compliance decision, not a prioritization problem. And if you need deep, qualitative interviews with your specific enterprise buyers, hire a researcher or consultant — the interview data is the expensive, irreplaceable input, and no platform generates it for you.
Independent studies on AI-assisted knowledge work — for example, the 2023 Harvard/BCG field experiment on consultants — found meaningful quality and speed gains on suitable tasks, alongside a "jagged frontier" where AI performs poorly on tasks outside its strengths. Roadmap synthesis sits inside that frontier; original customer discovery sits outside it. Use the tool accordingly.
What this looks like when the analysis is actually run
Five runs, five plans, and one of them has a deadline set by other people's contracts.
The subject is Verrano Pay, a sample company profile we use for testing rather than a customer: an SMB payments platform, $9.4B of annual volume, $84M net revenue, 28,000 merchants.
Excerpt from a real Percision run · Customer Value Architecture (T14) · sample company profile
First, because the window closes. Secure 3-year exclusivity on the three existing platform integrations — $1.8–2.4M over 18 months for 18–22×, or $42–48M of incremental lending revenue. Terminate if fewer than two platforms sign by Month 18, or if renegotiation windows do not materialise before December 31, 2026.
Second, because it protects the facility. A lending health-score dashboard issuing real-time covenant alerts before the 9.0% threshold is breached, with covenant headroom held at 120 bps or better — inside a $2.8–3.4M programme returning 2.1–2.8× within 24 months.
Third, because it needs the first two. Scaling the book from $110M to $260M at 22% take-up, taking lending from 22% to 38% of revenue.
Fourth, and most speculative. The marketplace: $2.1M–$4.2M over 36 months, gated on at least 2 capital providers committing by Month 12 and a 90-day rolling charge-off below 7.5%.
What the sequence is measured on. Exclusivity agreements: 3 existing plus 2 new platforms by Month 24. Near-zero-CAC acquisition share 75% by Month 36. Merchant 90-day retention lift of 4–6 percentage points by Month 12.
| Horizon | Projection |
|---|---|
| Year 1 | $2-3M incremental from deeper integration (12-month lag) |
| Year 2 | $12-15M incremental from exclusivity-protected lending origination |
| Year 3 | $28-30M incremental from two new platform integrations |
The exclusivity work goes first because its deadline is external. Renegotiation windows open when the contracts say they open, and a plan that misses December 2026 does not get a second attempt for three years — regardless of how attractive the other items are.
Everything downstream assumes the distribution holds. Scaling advances to $260M through platforms that could terminate on notice is a larger bet than the credit risk, which is why the cheapest item on the list is also the first.
Read a complete Percision report — every page, no email required.
FAQ
Q: How is JTBD different from just prioritizing by revenue impact? Revenue impact tells you what's valuable to you. JTBD tells you what's valuable to the customer — and underserved customer jobs are usually where durable revenue comes from. Use JTBD to find the bets, then apply revenue and cost filters as constraints.
Q: We're pre-product-market fit. Does JTBD still apply? Yes, more so. Before PMF, your entire job is finding one underserved job you can serve better than anyone. JTBD is a discovery tool as much as a prioritization one.
Q: Can Percision do the customer interviews for us? No. It structures and analyzes the inputs you bring. The interviews and workaround evidence are yours to gather — that's the part that can't be automated.
If you've done the discovery and want to turn a stack of jobs into a ranked, board-ready roadmap fast, you can try Percision here. Bring the customer evidence; let the platform pressure-test the sequencing.