What Should You Put on the Roadmap First in Healthtech? A Jobs to Be Done Approach
Direct answer: In healthtech, the first item on your roadmap should be the feature that lets your primary user complete their most important, most poorly-served job—not the feature with the loudest advocate or the flashiest demo. Use Jobs to Be Done (JTBD) to identify the underserved job with the highest stakes (clinical, financial, or workflow), then sequence the roadmap around progress on that job rather than around product categories. In healthtech specifically, "the job" almost always spans multiple stakeholders (clinician, patient, payer, admin), so the winning first move is the one that unblocks the whole chain, not just one persona.
Why healthtech roadmaps fail without JTBD
Digital health products serve users who did not ask for software. A physician wants fewer clicks and defensible documentation; a patient wants to feel less sick and less confused; a revenue-cycle manager wants clean claims. None of them wake up wanting "an app." When roadmaps are built around features ("add e-prescribing," "add a symptom checker"), teams optimize for parity with competitors instead of progress on a real job—and in a regulated, reimbursement-driven market, feature parity rarely wins deals.
Jobs to Be Done reframes the question. Instead of "what should we build," you ask "what is our user trying to make progress on, and what forces are holding them back?" Clayton Christensen's formulation is useful here: people "hire" a product to do a job. Your roadmap-first decision is choosing which job to help them do better than any alternative they currently hire—including the alternative of doing nothing.
Running JTBD in a digital health context
Here is a concrete walkthrough for a healthtech team deciding what ships first.
1. Name the job, not the feature. Write the job in the format "When [situation], I want to [motivation], so I can [expected outcome]." Example: "When a diabetic patient's readings spike between visits, the care team wants to intervene before an ER escalation, so they can avoid a preventable admission." Notice this job crosses clinician, patient, and payer value—that's typical in healthtech and it's a signal you're at the right altitude.
2. Map every current "hire." List what your users use today to do this job: phone tag, a spreadsheet, an EHR flag, a rival product, or nothing. For each, note the friction. In healthtech, "nothing" is a common competitor—the ER admission that no one prevented. Underserved jobs with expensive non-consumption are your strongest roadmap candidates.
3. Score jobs on importance × satisfaction. A good JTBD analysis rates each candidate job on how important it is to the user and how satisfied they are with current solutions. High importance + low satisfaction = your roadmap headline. Ask directly: "How much does completing this job matter?" and "How well do current tools do it?" Don't guess—pull from customer interviews, support tickets, and, where you have them, claims or utilization patterns.
4. Test the outcome measures. For each candidate job, define what "good" looks like in the user's terms—not yours. Clinicians measure success in time saved and risk reduced; payers in avoided cost and quality scores; patients in confidence and reduced burden. If you can't state the outcome metric, you don't understand the job yet.
5. Sequence by unblocking, not by demand. The first roadmap item should remove the single biggest force preventing progress on the highest-value job. Sometimes that's a headline feature; often it's an unglamorous enabler (an integration, an eligibility check, a consent flow). In healthtech, a beautiful clinical feature that can't push data into the EHR does zero jobs.
What "good" looks like: a one-page roadmap thesis stating the job, the target user, the outcome metric, the current alternatives you beat, and the first shippable increment that produces measurable progress on that job.
Where Percision fits—and where it doesn't
Full disclosure: I write for Percision, an AI-powered strategic intelligence platform (percision.app), so treat this as one option among several.
Percision includes Jobs to Be Done among its 27+ frameworks. For a healthtech team, it's most useful after you've done the qualitative discovery. You feed in your business context—users, current alternatives, candidate jobs, market and reimbursement dynamics—and it runs that through structured reasoning steps to produce a prioritized job map, a roadmap thesis, and a board-ready deck you can take into a planning cycle. Because it also runs financial intelligence (DCF, scenario analysis, ratios), you can pressure-test whether the highest-value job is also the one with a defensible business case—useful when a payer-facing job and a patient-facing job compete for the same quarter.
It's positioned as a co-pilot, not an autopilot: it structures and accelerates the analysis; your leadership team owns the judgment. BCG and Harvard Business School researchers have documented meaningful productivity gains when knowledge workers use generative AI on well-scoped tasks—but the same research warns of a "jagged frontier" where AI confidently gets things wrong. Roadmap prioritization sits near that edge, so keep humans on the clinical and regulatory calls.
When you don't need Percision. If you have three candidate jobs and strong customer-interview data, a whiteboard and an importance/satisfaction 2×2 will get you there in an afternoon. If your bottleneck is talking to users—not synthesizing what they said—no platform substitutes for that. And for FDA pathway, HIPAA, or clinical-validation questions, hire a domain human, not a general strategy tool.
You can run a JTBD-driven roadmap analysis at percision.app if you want the structured version.
FAQ
Q: In healthtech, whose job do we prioritize when clinician, patient, and payer all matter? Prioritize the job whose completion unblocks the others. Often that's the payer or workflow job, because reimbursement and adoption gate everything else—but validate with your specific buyers.
Q: How is JTBD different from user stories? User stories describe features from a role's view ("As a nurse, I want..."). JTBD describes the underlying progress a user seeks regardless of solution, which keeps you from over-indexing on one persona's requested feature.
Q: Can we skip customer interviews and just use AI to generate the job map? No. AI can structure and prioritize jobs, but the raw signal—what users actually struggle with—has to come from real conversations, tickets, and data. Garbage in, confident garbage out.
Disclosure: This article was written by Percision's content team. Percision is one strategy tool among many; choose what fits your stage and budget.