TL;DR Market adoption isn't proven by a working prototype, a strong patent, or a few encouraging investor meetings. It's proven when a specific buyer changes their behavior — pays for it, switches to it, and folds it into how they already work — because it solves a problem they were already trying to solve. This article is about telling product validation from market validation before you spend years and millions building the wrong thing.
A founders story
In my recent conversation with Nathalya Mamane and Dr. Michael Mina, co-founders of Anywhere Dx, one idea kept coming back. Successful healthcare innovation starts by understanding the customer's problem deeply, not by asking what the technology can do.
Nathalya didn't begin as a diagnostics scientist. She began as a mother who was tired of losing a full day every time one of her kids needed a strep test.
That frustration, not a molecular assay, was the start of the product.
I think about that story a lot, because it runs against how most medical device and health tech companies begin. The usual pattern is to take something that already works in a clinic, make it smaller, and assume families or patients will adopt it because the science is better. Anywhere Dx started from the other end — with the question, "what would a parent actually want to use?" — and then went looking for the technology that could deliver it.
The difference sounds small. It decides everything.
Why do great products still fail to get adopted?
This is one of the most common patterns I see working with founders.
A company develops breakthrough technology. The engineering team solves genuinely hard problems, the bench testing looks great, and the intellectual property is strong. Everyone involved starts to assume the hard part is behind them.
Then the product launches.
Sales are slower than expected. Pilot programs don't convert into paying customers, and investors begin asking uncomfortable questions. Marketing gets blamed. Sales gets blamed. Pricing gets blamed.
The website gets redesigned. New campaigns launch. More money goes into lead generation.
None of those things address the real issue. The company never proved the market wanted the solution in the first place.
There's a distinction founders miss here, and it's worth saying plainly. You can prove your product works. That doesn't prove people will buy it, and it definitely doesn't prove they'll change what they do today to make room for it.
Proving the technology works is an engineering question. Proving the market will adopt it is a behavior question. They live in different buildings, and the second one can be harder.
What's the difference between technical validation and market validation?
Technical validation answers: does it work, reliably, within spec? That's your engineers, your bench data, your regulatory pathway, your quality system. It's necessary. It's also the part most technical founders are naturally good at, which is exactly why it gets over-weighted.
Market validation answers a different question: will a specific person, with a specific job to get done, change their current behavior to use this — and keep using it?
Here's why the second one gets skipped. It's uncomfortable. It asks you to go find out whether people actually care, and the answer might be "not enough to switch." When the prototype is humming and the demos land well, it's much easier to treat enthusiasm in the room as evidence of demand in the market.
It isn't. A great demo tells you the product is impressive. It tells you nothing about whether someone will reorganize their day, their budget, or their clinical workflow around it.
The clearest way I know to keep these straight is Clayton Christensen's Jobs-to-Be-Done framing. In his well-known study of why people bought milkshakes, the winning question wasn't "how do we make a better milkshake?" It was "what job is the customer hiring this milkshake to do?" (Harvard Business School Working Knowledge tells the story in full.) Adoption happens when your product is the best available candidate for a job the buyer is already trying to get done. If there's no job, there's no adoption, no matter how good the assay is.
I've seen the tech-first version of this play out more than once. A team builds a genuinely elegant device around a capability they're proud of, gets strong bench data, raises on the strength of it, and only starts asking "who exactly is this for, and what will they stop doing to use it?" after the first pilots go quiet. The science was never the problem. The order was. They validated the technology thoroughly and the market barely at all, then spent the next year trying to retrofit a job onto a product that was already finished.
Anywhere Dx ran the sequence in reverse, and that's the whole point of Nathalya's story. The job came first. The molecular chemistry came second, in service of the job. When you build in that order, market validation isn't a gate you hit at the end. It's the thing steering every requirement along the way.
Why does every month building the wrong thing cost more than the last?
Because the cost of being wrong compounds.
Every month spent building around an assumption instead of evidence adds weight to that assumption. Features get added. Patents get filed. Regulatory work begins. Manufacturing decisions get locked in.
Then the launch comes, and the uncomfortable question finally gets asked out loud: "why isn't anyone buying?"
By that point, changing direction is expensive in a way it never was at the start. The design is committed. The submission is written around the current claims. The team has spent two years telling investors a specific story. Reversing any one of those has real cost, and reversing all of them at once can end the company.
I've watched teams spend years refining a product around a belief nobody ever tested. The refinement is real work, and it feels like progress. It's just progress in a direction the market never asked for.
The good news is that testing the assumption early is cheap by comparison. A month of honest customer discovery costs a rounding error against a year of engineering. The problem is rarely budget. It's that discovery feels less like progress than building does, so it keeps getting deferred.
What does real market validation actually look like?
Adoption starts long before launch, and it's built from evidence you can gather now. A few areas matter more than the rest.
Customer discovery that goes past the interview. Interviews are a start, but people are unreliable narrators of their own future behavior. Watch what they do, not only what they say. Ask what they've already tried, what they're paying for today, and what they gave up on. Past behavior predicts adoption far better than stated intent.
The job, defined precisely. Nathalya's job wasn't "run a molecular test." It was "confirm whether my kid has strep without losing a day to the pediatrician and the pharmacy." Define the job at that level and the product requirements almost write themselves.
Who actually buys, and who has to say yes. In healthcare the person who uses the product often isn't the one who pays, prescribes, reimburses, or approves it. A parent may want an at-home strep test, but adoption at scale can still run through pediatricians, payers, and regulators. Map every person who can say no. Any one of them can hold up the whole thing.
Workflow fit. A product that forces someone to change five things about their day will lose to a worse product that fits how they already work. Friction is what kills adoption, and it rarely shows up in a demo. Find every step where your product asks the user to do something new and ask whether they truly will.
Clinical value and economic value are not the same. A device can be clinically better and still fail commercially because no one will pay for the improvement. Separate the two questions early: is it better, and will someone open their wallet for "better"? You need both answers, and they come from different people.
Leading indicators of adoption. Long before revenue, watch for the small signals that someone has genuinely changed behavior: they use it without being reminded, they ask how to buy more, they tell a colleague, they get annoyed when it's taken away. Those beat any survey score.
None of this requires a finished product. Most of it you can learn with a rough prototype, a clear description of the job, and the willingness to hear "no."
Why is buyer indecision the adoption risk founders underestimate most?
Most founders picture the adoption battle as their product against a competitor. In complex, regulated markets, that's not usually where deals die.
Matthew Dixon and Ted McKenna analyzed more than 2.5 million sales conversations for The JOLT Effect and found that 40 to 60 percent of qualified deals are lost not to a competitor, but to "no decision" — the buyer, interested and qualified, simply doesn't act (Dixon discusses the finding here). Most of that loss comes from indecision and fear of getting the choice wrong, not from preferring the status quo.
Sit with that for a second, because it reframes what "market adoption" even means.
Your real competition often isn't the other device. It's the buyer's decision to do nothing — to wait, to keep coping with the current mess, to avoid the risk of championing something new inside a cautious organization. A better product doesn't beat inertia on its own. You beat inertia by making the decision to adopt feel safe, evidenced, and low-regret. That's a commercial job, and it starts in how you design and position the product, not in the sales deck at the end.
What are the most common ways founders misread adoption signals?
When teams believe they've validated the market and haven't, the mistake usually falls into one of a few familiar traps. They're worth naming, because each one feels like evidence at the time.
Treating enthusiasm as intent. A prospect who loves the demo and a prospect who reorganizes their budget around you are not the same person. Excitement is free. Adoption costs the buyer something.
Counting pilots as adoption. A pilot proves someone is willing to try you at low risk. It proves nothing about whether they'll pay, renew, or roll you out. Watch what happens after the pilot ends and the free help goes away.
Listening to the user and ignoring the buyer. In healthcare the person who wants your product often can't approve it. If you've only validated the enthusiastic user and never the person who signs, you've validated half the decision.
Confusing "better" with "worth switching for." Improvement doesn't create adoption. The improvement has to clear the very real cost of change — retraining, revalidation, new workflows, internal risk. Small gains rarely clear that bar.
Reading your own conviction as market signal. The more you've invested, the harder it is to hear "no." Founders often mistake their own certainty for demand, then interpret slow sales as a messaging problem.
The common thread is that all of these feel like progress. That's what makes them dangerous. Real validation is uncomfortable on purpose, because it's actively trying to find the reason someone won't adopt, while these traps are all busy confirming that they will.
A useful habit: for every piece of "proof" you're excited about, ask what it would look like if the opposite were true. If a great pilot and a pilot that went nowhere would produce the same conversations in your building, the pilot isn't telling you what you think it is.
Where a strategic growth advisor earns their keep
Most founders in this position don't need another marketing agency. They need someone who will pressure-test the commercial assumptions before those assumptions harden into expensive engineering decisions.
That's most of what I do as a fractional CMO and strategic growth advisor. Not "how do we get more leads." Earlier questions that decide whether leads will ever matter:
Are we validating a product, or validating a market? They're different, and only one predicts revenue.
Whose behavior has to change for this to work, and how big a change are we asking for?
Who has to say yes besides the user, and what does each of them need to see?
Is there real economic value here, or only clinical value nobody will pay for?
What would have to be true for this to fail, and how cheaply can we test that right now?
The value of an outside partner here is objectivity. When you've spent two years and real money building something, you lose the ability to hear "no." A good advisor's job is partly to protect you from your own conviction — to connect customer research, positioning, and workflow reality before they get locked into a roadmap you can't easily unwind.
Done early, that work is cheap insurance. Done late, it's a rescue.
The founder takeaway
Your goal was never to prove your technology works. That's table stakes, and you were always going to clear it.
Your goal is to prove that a specific person will change their behavior because of what you built — that there's a real job, a real buyer, and a real reason to act now instead of later.
Anywhere Dx understood that from day one. They didn't start with a molecular platform looking for a use. They started with a frustrated parent and worked backward to the technology. That's the order that produces products people adopt instead of products people admire.
Frequently asked questions
What's the difference between product validation and market validation?
Product validation proves the technology works reliably. Market validation proves a specific buyer will change their behavior and pay to use it. You can pass the first and fail the second, and the second is what predicts revenue.
How early should we test market adoption?
Before you lock design, regulatory claims, and manufacturing. Customer discovery with a rough prototype costs a fraction of a year of engineering, and it's the cheapest point at which you can still change direction.
Aren't customer interviews enough to prove demand?
No. People are poor predictors of their own future behavior. Weight what they already do, pay for, and have abandoned over what they say they'll do — behavior predicts adoption better than intent.
Why do clinically superior devices still fail commercially?
Because clinical value and economic value are separate questions. A device can be measurably better and still fail if no one — user, payer, or system — will pay for the improvement or change their workflow to get it.
Our demos go great. Isn't that a good sign?
It's a sign the product is impressive, which is different from adoptable. Enthusiasm in the room doesn't tell you whether someone will reorganize their budget or workflow around it. Watch for behavior change, not applause.
What are the earliest signs the market will actually adopt?
Unprompted repeat use, requests to buy more, referrals to colleagues, and visible frustration when the product is taken away. These leading indicators beat any survey score.
Where to start
If you're not sure whether you're validating a product or validating a market, that's the thing to settle before you spend another six months building. It's a short, honest conversation, and it's a lot cheaper than finding out at launch.
Bring me the messy version. We'll figure out which one you're actually proving.
