Product-market fit (PMF) is the point at which a product satisfies a market that has enough real buyers to pull the product out of the company. Before a founder can test for it, there has to be something to test: a written account of which problem the company solves, for whom, and why it can become a business. That account is the foundation. Without it, every pilot tests a guess. Building it starts with knowing how to find a problem worth solving for a healthcare startup.

The evidence points the same way. In CB Insights’ review of 111 startup post-mortems, 35% of founders named “no market need” as a reason their company failed. More of them, 38%, named running out of cash. In A|P’s reading, cash is usually where the story ends rather than why it ends: a company solving a problem that buyers will pay to have solved tends to find the money to keep going.

Healthcare adds a second filter. The doctor you interview is less and less often the person who buys. Consider for instance, in 2024, 42.2% of US physicians worked in a practice owned wholly by physicians, down from 60.1% in 2012, and 34.5% worked in a hospital-owned practice. For a growing share of the clinicians you will meet, the purchase runs through a health system’s finance team, its product review committee and its IT security review.

This guide is written for founders of healthtech companies who are about to decide what to build. It also serves founders who built first and now have unpaid pilots that won’t convert: for them, the foundation is the step they skipped, and the place to go back to. It covers how to find a problem worth solving for a healthcare startup, and how to write the other five parts of the foundation so they hold together.

The A|P series on PMF foundations

This page is the hub for A|P’s series on the foundation of product-market fit. Each article takes one part of the foundation further.

• Societal need isn’t a market: what makes a healthcare market worth entering (publishing soon)
• Solution looking for a problem: from an initial insight to a problem statement that passes four tests (publishing soon)
• The doctor loves it, the CFO won’t sign: picking your first customer segment (publishing soon)

The foundation sits inside a longer sequence. Validating it, measuring whether you have reached fit and building traction are covered on A|P’s page on product-market fit for healthtech ventures.

What is product-market fit, and why does the market come first in healthtech?

Product-market fit means a product satisfies a good market: one with many real buyers who feel an acute problem now, hold a budget and have the authority to spend it. The market matters more than the product, and the product matters more than the team. A strong team in a weak market still loses.

Since the days Andreessen wrote about it, a good market is easy to recognise from the inside. Buyers pull the product out of the company. The product only has to work at a basic level, because buyers who need it badly will tolerate rough edges. A weak market feels different. Sales cycles drag. Prospects ask for custom builds before they sign. Meetings end with agreement that the problem is real, and no contract follows. Every closed deal feels like hand-to-hand combat.

Healthcare makes one confusion especially easy: mistaking a societal need for a commercial market. Clinician burnout, documentation backlogs and avoidable readmissions are real and large. A market exists only where someone has a budget, the authority to spend it and a reason to spend it this year. If the physician loves the app but the health system’s CFO won’t approve the purchase, there is no market yet, only a cause. Hospital finance leaders are setting that bar high: in a January 2026 survey of 150 US hospital CFOs commissioned by the vendor SmartSense, 51% said new technology must return more than 110% within 18 months.

One part of the classic advice has aged. When software was expensive to build, a product that basically worked could stand alone for years. It no longer can. KLAS counted 657 AI vendors in use or under consideration at US healthcare organisations in 2025, and more than half were mentioned only once. A working product is now the entry ticket, not the advantage. Distribution, workflow integration and defensibility decide who keeps the market.

The practical signal is simple. If you are extending free pilots, rewriting your pitch deck, tweaking your website, and chasing the same prospect for the sixth time, you do not have product-market fit yet.

Why write the foundation before building an MVP?

Writing the foundation first lets a founder test the cheapest assumptions before spending runway on the most expensive one. Build-first methods treat a minimum viable product (MVP), the smallest version of a product in which every essential part works to a defined standard, as the main test of an idea. In healthcare that test is slow, costly and hard to read.

Building first has four gaps:

1. It undervalues planning. A plan can be thoughtful without being precise, and a written plan exposes contradictions that code hides.
2. It treats shipping as the only test. Interviews, expert advice, market research and pre-sales can test most assumptions before an MVP exists.
3. It never says how much evidence is enough. Iterating on data is sound, but without a written hypothesis there is nothing to judge the data against.
4. It assumes traction follows a good product. Good products still fail without enough buyers. Product and traction have to be built together.

The cost of a failed test is higher in healthcare than in most markets. A pilot with a health system can take months to set up, and it often ends without a clear answer. In the same SmartSense survey, 57% of hospital CFOs said half or more of their technology pilots fail. A seed-stage company with nine months of runway can afford very few of those.

The foundation has six parts. Each answers one question, and each has to be written twice in a B2B healthcare business: once for the person who decides and pays, and once for the person who uses the product every day.

1. Problem to solve: what problem are you solving?
2. Target audience: who has it badly enough to pay?
3. Value proposition: how will you solve it, in terms of benefits rather than features?
4. Competitive advantage: why are you the one who can solve it, now and later?
5. Growth strategy: how will you reach more of these buyers over time?
6. Business model: how does all of this become revenue that exceeds cost?

The parts are written in prose, not bullets. A paragraph forces cause and effect: if the problem doesn’t lead to the audience, and the audience doesn’t lead to the value proposition, the gap shows on the page. Every part rests on one starting point, the initial insight.

Where does a good initial insight come from?

A good initial insight is an observation about a market that is earned, unique and grounded. Earned means it comes from experience, not a brainstorm. Unique means it is not what everyone already believes. Grounded means it points to at least one of the six parts of the foundation. An insight that fails any of the three makes a weak foundation.

Each quality has a test:

• Earned. The insight comes from one of three places: your own professional or personal experience; behaviour you watched users show with an existing product or a workaround; or long immersion in the problem, through many conversations with buyers and hands-on use of the tools they already have. Ideas generated in a room, away from the work, rarely pass.
• Unique. Share it and watch the reaction. If everyone agrees, it is an obvious truth and the market will be crowded. If nobody agrees, it is probably wrong. Mixed reactions, with some people persuaded, are the signal to keep going.
• Grounded. The insight has to change at least one part of the foundation: who the buyer is, what the product must do, why the company can win, or how it makes money. An insight that changes none of them is an opinion, not a starting point.

An illustrative example from US healthcare: a founder spent six years as a clinical documentation specialist at a Midwest health system, reviewing physician notes before they went to coding. She saw that the system judged its AI scribe pilot on minutes saved per physician, but its finance team cared about something else: diagnoses the physician discussed but never documented, which meant care that was delivered but never billed. Her insight: a documentation tool sold on time saved competes with every scribe on the market and with the electronic health record (EHR) vendor’s own product; a tool sold on documented, billable care answers the CFO’s question instead of the physician’s.

The insight is earned (six years inside the workflow), unique (the market talks about clinician time), and grounded in two parts of the foundation: the business model and the competitive advantage. Whether it is right is a question for validation. It is a sound place to start.

How do you find a problem worth solving for a healthcare startup?

To find a problem worth solving for a healthcare startup, start with the outcome your buyer is trying to reach, ask why they want it, and then list what stops them from getting it today. Those obstacles are your problem statement. Then test the statement for four common faults before you build anything.

The most common failure at this step is a problem that isn’t acute enough. The product solves something real that nobody feels strongly enough to pay to have solved. It usually happens because the founder started from a solution and wrote a problem to fit it.

Three questions keep the problem separate from the solution:

1. Outcome: what is the customer trying to achieve? Describe the result they want, not the product you imagine. A primary care group wants to see more patients without adding physicians.
2. Motivation: why do they want that outcome? Stay with the immediate economic or operational reason, such as revenue per physician, wait times or retention of staff. Deeper motives like status are real, but no product can close a gap called “lack of status”.
3. Gap: what stops them today? Physicians spend evenings finishing notes, so adding appointment slots means losing physicians. Each gap you write down is a hypothesis for the problem you will solve.

Then run the statement through four tests:

1. The active-pain test. Read the first verb. If the statement starts with a builder’s verb (helping, enabling, improving, providing), it is a solution in disguise. “Helping physicians finish notes faster” fails. “Physicians finish notes after clinic hours” passes, because it describes the customer’s friction without mentioning your product.
2. The severity test. Ask whether solving the problem perfectly would unlock the customer’s motivation, or only remove an annoyance. “Physicians dislike typing notes” is a minor irritation. “The practice can’t add appointment slots without losing physicians” is the problem a buyer pays to remove.
3. The cause-and-effect test. Write the outcome, motivation and gaps as one paragraph and read it aloud. The outcome should lead to the motivation, and the gaps should block it. If the sentences don’t follow from each other, the statement doesn’t hold together yet.
4. The solution-independence test. Ask whether you could validate the statement with customers before writing a line of code. “Practices want shorter documentation time but can’t get it because they don’t have an AI scribe” fails: the product has taken the place of the gap.

A problem statement that passes all four is ready for validation. One that fails any of them will look like customer interest in interviews and turn into stalled pilots later.

Who should your first customer segment be?

Your first customer segment should be the smallest group that feels the problem most acutely and can pay to solve it. Win that group, expand to adjacent segments, and repeat. Starting narrow gives the company a target it can actually reach and brings revenue in before runway runs out.

The most common failure at this step is a target audience that is either too broad to reach or too small to sustain the company. A good segment comes from the bottom up. By the time you have an insight and a problem statement, your market is no longer every US provider; it is only the buyers who experience those specific gaps.

Build it in three steps:

1. List the attributes that matter. Practice type and ownership, specialty, size, EHR, payer mix, and the job title of the person who would sign.
2. Combine them into segments. For example: physician-owned ophthalmology groups with fewer than ten physicians; hospital-owned primary care practices in a regional health system.
3. Sort the segments into “now” and “future”. A “now” segment must have at least one of five traits: the most acute problem, the strongest response to your value proposition, the clearest advantage over competitors, the easiest reach, or the strongest willingness to pay.

Every segment in a B2B healthcare business has two roles, and they rarely want the same thing. The decision maker approves the budget and judges the business case. The end user works with the product every day. Write the problem and the value for each separately.

Ownership decides who those people are. In a physician-owned practice, the physician is often buyer, champion and user at once, and a sale can close in a meeting or two. In a hospital-owned practice, the physician who wants your product hands you to a value analysis or product review committee that evaluates new products before purchase, and to an IT security review. Any product that touches patient records also needs a business associate agreement under HIPAA, the US federal law that protects patients’ health information: under HHS guidance, a software vendor that needs access to a provider’s protected health information (PHI) must sign one before it gets access. The mix varies widely by specialty: in 2024, 70.4% of ophthalmologists but only 30.7% of cardiologists worked in private practices.

One caution applies to every “now” segment chosen because competitors ignore it. Ask why they ignore it. If the answer is that the segment can’t pay, can’t be reached or won’t stay, the gap in the market is not an opportunity. Choose a segment that is underserved and can also fund your next eighteen months.

How do you turn the problem into a value proposition?

A value proposition states what the customer gets, in their terms, not what the product does. The quickest way to write one is to draft the homepage you would want: a one-line promise at the top, then three to five distinct benefits, each tied to a gap in your problem statement. If a benefit can’t be traced to a gap, it is a feature looking for a reason.

The most common failure at this step is a value proposition that doesn’t land with buyers, usually because it lists features. Customers rarely describe their problems in the vocabulary of a feature list, so a features-first proposition gives interviews nothing to react to.

For the documentation founder above, the promise might read “Every diagnosis you discuss, documented and billed,” followed by benefits such as notes finished before the physician leaves clinic, and documentation the coding team doesn’t have to send back. Features come later, once interviews confirm the benefits matter.

What makes a healthtech advantage last?

A healthtech advantage lasts when it rests on something competitors cannot copy quickly: high switching costs, a resource only you can access, or a business model incumbents won’t adopt. A head start is not an advantage. Plan for two horizons: how you win the first customers, and what stops others from copying you once you do.

The most common failure at this step is a product that isn’t meaningfully different from the alternatives, which turns early fit into a price fight.

The short term: an advantage for whom? In the short term, what matters is not what your advantage is but whom it serves. Map three groups of competitors: direct competitors with a similar promise; indirect alternatives that solve the same problem differently, such as human scribes or offshore transcription; and adjacent markets that don’t compete today but could. Then find the segment the others serve badly. Adjacent markets move fastest in healthcare. When Epic launched its own ambient speech tool in August 2025, most of the 35 Epic customers KLAS interviewed expected to adopt it within two years.

The long term: one of the 7 Powers. A lasting advantage maps to at least one of the 7 Powers: scale economies, network economies, counter-positioning, switching costs, branding, cornered resource and process power. Three tend to matter most in US healthtech:

• Switching costs. A product built into the EHR workflow, billing rules and staff routines costs a provider months to replace.
• Cornered resource. Preferential access to something valuable, such as an exclusive data partnership or a clinical network, that a new entrant cannot buy on the same terms.
• Counter-positioning. A business model an incumbent won’t copy because it would damage its existing revenue, such as pricing on outcomes when the incumbent charges per seat.

None of these needs to exist on day one. All of them take longer to build than founders expect, so the foundation should name which one the company is working towards.

How will you grow, and will it pay?

A healthtech startup needs two growth plans and one business equation. The first plan is a traction channel that gets the first ten paying customers, even if it can’t scale. The second is a growth channel that can reach a large share of the market later. The business equation shows how revenue, pricing, customer lifetime value and costs combine into profit.

The most common failures at this step are a company with no sustainable way to grow, and a business model considered only at the surface.

Traction and growth are different channels. Early traction usually comes from things that don’t scale: founder-led sales to physician-owned practices, introductions through clinical advisers, specialty society meetings. Long-term growth comes from repeatable loops such as a sales team, channel partners, EHR vendor marketplaces or content that brings buyers in. Two traps sit on either side. Scaling paid acquisition before customers stay fills a leaking bucket. Staying on manual tactics too long keeps growth expensive and flat. How to build both channels is the subject of A|P’s page on go-to-market strategy for healthtech ventures.

The business equation has four levers.

1. Revenue streams: actual sources of cash, such as subscriptions, per-transaction fees or a share of savings. An audience is not a revenue stream.
2. Pricing: the payment model and the buyer’s real budget. It must match what the buyer is paid or measured on.
3. Lifetime value (LTV): the revenue a typical customer brings over the whole relationship.
4. Cost structure: the large fixed and variable costs of winning and serving customers, including customer acquisition cost (CAC), the average cost of winning one new customer.

Pricing is where US healthcare punishes a surface-level model. A practice running remote patient monitoring gets paid by Medicare only for patients whose devices send data on at least 16 days in a 30-day period. A vendor that charges per device shipped asks the practice to pay for patients it can’t bill for. A vendor that charges per patient who meets the threshold has a price the practice can defend to its own finance team.

How to write your PMF foundation: a practical test

You can draft the whole foundation in an afternoon. Write each step in full sentences.

1. Write your initial insight in two sentences, and name where it came from: your experience, behaviour you observed, or time spent immersed in the problem.
2. Write the outcome, motivation and gaps as one paragraph. Run it through the four tests: active pain, severity, cause and effect, solution independence.
3. List the attributes that define your buyers, combine them into segments, and choose one “now” segment. Name its decision maker and its end user.
4. Write a homepage promise and three benefits. Check that each benefit traces back to a gap.
5. Name one direct competitor, one indirect alternative and one adjacent player. Name the one Power you are building towards.
6. Name one traction channel and one growth channel. Write the four levers of your business equation.
7. Read all six parts aloud as one piece of prose. Mark the sentence you are least sure of. That is the first assumption to validate.

Frequently asked questions

What is a good product-market fit framework for healthcare startups?

A good product-market fit framework for healthcare startups covers six parts: the problem, the target audience, the value proposition, the competitive advantage, the growth strategy and the business model. In B2B healthcare, each part has to be written for both the person who buys and the person who uses the product, because in US healthcare they are often different people.

Why do healthcare startups fail?

Healthcare startups most often fail because they solve a problem that buyers won’t pay to have solved. In CB Insights’ review of 111 post-mortems, 35% of founders cited no market need and 38% cited running out of cash. In US healthcare the risk is sharper, because the clinician who likes a product is often not the person who approves the purchase.

Do I need an MVP before I write my PMF foundation?

No. The foundation comes before the MVP, because it tells you which assumptions to test first. Many of them, such as whether the problem is acute or who signs the contract, can be tested through interviews and pre-sales before any product exists.

What if doctors love the product but the hospital won’t buy it?

If doctors love the product but the hospital won’t buy it, you have found a user, not a buyer. Rewrite the problem and the value proposition for the decision maker, usually finance, operations or the product review committee, in terms of what they are measured on. Or choose a first segment where the physician is also the buyer, such as a physician-owned practice.

How long does it take to find product-market fit in healthcare?

There is no fixed timeline, and runway sets the real limit. Healthcare sales and pilots are slow, so a seed-stage company can run only a few full tests before its cash runs out. Writing the foundation first makes each test count.