The Search You Didn't Know You Signed Up For

There is a story first-time founders tell themselves, usually without realizing they’re telling it. It goes like this: I will build the product, I will launch the product, people will discover the product, and the company will begin. The story has a satisfying shape. It has a beginning, a middle, and a launch day. It is also, in almost every case, fiction. What actually happens is stranger and slower. You build the product. You launch it. A modest number of people look at it, a smaller number try it, and a much smaller number stick around. The metrics don’t say “yes” and they don’t say “no.” They say something closer to “hmm.” And then you enter a phase of company-building that nobody adequately warned you about: a long, ambiguous, often demoralizing period of adjusting what you’ve built and how you talk about it, over and over, until something clicks. Or doesn’t. This essay is about that period. It’s about two ideas that sound like jargon but describe the most important work you will do as a founder: product/market fit and message/market fit. If you’ve never worked in go-to-market roles before, which is to say sales, marketing, or anything else concerned with how a company finds and wins customers, these terms may be new to you. That’s fine. The concepts are simple to define and brutally hard to achieve, and the gap between those two facts is where most startups die. \ What product/market fit actually means The term “product/market fit” was popularized by Marc Andreessen in a 2007 blog post, building on thinking from his Stanford professor and benchmark investor Andy Rachleff. Andreessen’s definition is almost anticlimactic: product/market fit means being in a good market with a product that can satisfy that market. Let’s unpack that, because every word is doing work. A “market” is not “everyone who might conceivably use this.” A market is a group of people or companies who share a problem, know they have it, and are willing to spend money or change their behavior to solve it. That last clause matters most. Plenty of problems are real but not urgent. People will agree, in conversation, that yes, that’s annoying, that’s inefficient, somebody should fix that. Then they will go back to their day and never think about it again. A real market is made of people for whom the problem is painful enough that they are already trying to solve it, badly, with spreadsheets and duct tape and workarounds. The existence of ugly workarounds is one of the most reliable signals of a real market because it proves people care enough to suffer for a solution. A product that “satisfies” that market is one that solves the problem well enough that people adopt it, keep using it, and would be upset if you took it away. Not impressed by it. Not complimentary about it. Dependent on it. Product/market fit, then, is the state where these two things connect: a genuine, motivated demand and a product that genuinely answers it. When you have it, you can feel it. Customers pull the product out of your hands. Word of mouth starts happening without you. Your problem shifts from “how do we get anyone to care” to “how do we keep up?” Rachleff liked to say that you can always feel when product/market fit isn’t happening, and you can always feel when it is, and the difference is unmistakable. The corollary, which founders resist, is that if you’re not sure whether you have product/market fit, you don’t have it. \ What message/market fit means, and why nobody tells you about it Product/market fit gets all the attention. Message/market fit is its quieter sibling, and ignoring it is one of the most common ways technical founders sabotage perfectly good products. Message/market fit means you have found the words. Specifically: you have found a way to describe what your product does, for whom, and why it matters, such that the right people hear it and immediately understand that it’s for them. The message lands. It doesn’t require explanation or a demo or a twenty-minute call to make sense. The prospect reads your home page or hears your one-liner, and something in their head goes: oh. That’s my problem. That’s me. Here’s why this matters as much as the product itself: your product cannot speak for itself. This is one of the harder truths of early-stage company building. You believe, deeply, that if people would just try the product, they would get it. You’re probably right. But “just try the product” is itself a sale, and the only tool you have to make that sale is language. Before anyone experiences what you built, they experience how you describe what you built. If the description fails, the product never gets its chance. A great product with a bad message loses to a decent product with a great message far more often than engineers want to believe. And here is the part that surprises first-time founders most: you, the person who understands the product best, are usually the worst-positioned person to find the message. You know too much. You describe the product in terms of what it is and how it works because that’s how you think about it, having built it. The customer doesn’t care what it is. The customer cares what it does for them, in their language, against the backdrop of their existing frustrations. The message that works is almost never the message you would have written on day one, and you cannot derive it from first principles. You have to find it the same way you find product/market fit: by trying things in front of real people and watching what happens. \ Two searches, one loop It would be convenient if these were sequential, if you could nail the product first and then figure out the messaging. They’re not, and you can’t, because each one is the instrument you use to measure the other. Think about what happens when you put your product in front of a prospect and they don’t bite. What went wrong? Maybe the product doesn’t solve a problem they care about. Maybe it solves a problem they care about, but your description of it was so off-key that they never realized it. Maybe the product and message are both fine and this just wasn’t the right person. From the outside, all three failures look identical: a shrug, a polite “interesting,” a trial that goes nowhere. The signal doesn’t come labeled. This is why early-stage iteration is so disorienting. You are running an experiment with at least three free variables, the product, the message, and the audience, and every result is ambiguous about which variable caused it. A scientist would call this a confounded experiment and redesign it. A founder doesn’t have that luxury. You can’t isolate variables when each test takes weeks and your runway is measured in months. So you do the unscientific thing: you form a judgment about which variable is most suspect, change it, and run the loop again. The loop itself is simple to describe. You make a hypothesis: these people have this problem, our product solves it, and these words will make that obvious. You expose the hypothesis to reality through sales calls, a landing page, cold outreach, a demo, and a launch. You collect the signal, which is usually faint and contaminated with noise. You revise the hypothesis. You go again. What nobody tells you is how many times you’ll run that loop. Not three. Not five. Often dozens, across a year or two, with most iterations producing results that are slightly worse than ambiguous. The companies whose origin stories you’ve read have compressed this period into a sentence, usually something like “after some early experimentation, we found our footing.” That sentence is hiding eighteen months of someone’s life. \ The signals that lie to you Part of what makes the search slow is that the early evidence actively misleads you in both directions. It’s worth cataloging the major lies because you will encounter every one of them. The first lie is enthusiasm without commitment. People are nice. When you show someone your product, especially someone who knows you, they will say encouraging things. “This is really cool.” “I could see a lot of people using this.” “Send me a link when it launches.” None of these statements cost the speaker anything, and statements that cost nothing carry no information. The only signals that matter are the ones with a price attached: money, time, reputation, switching cost. Did they pay? Did they integrate it into their workflow? Did they introduce you to a colleague, putting their own credibility behind you? Did they come back next week without being prompted? Compliments are noise. Behavior is a signal. The second lie is the wrong customers saying yes. Sometimes you’ll get traction, real, paying traction, from people outside the market you intended to serve. This feels like success and might be. But it might also be a trap: a small pocket of demand that doesn’t generalize, pulling your roadmap toward customizations that serve five customers and repel five thousand. Early money is intoxicating precisely because you have so little of it, and it takes discipline to ask whether the people buying represent a market or merely a handful of anomalies. The third lie is the false negative. You pitch the product to twenty people, nobody converts, and you conclude the product is wrong. But maybe the product was right and the message was wrong, or the message was right and those twenty people were the wrong audience. Promising companies have been abandoned over results that condemned the words, not the work. This is why the entanglement of product fit and message fit is so dangerous: a failure in either one can masquerade as a failure of the whole. The fourth lie is the metric mirage. Signups that don’t activate. Traffic that doesn’t convert. A launch-day spike from a community that came for the novelty and never returned. Vanity metrics are called that because they flatter you, and flattery, again, is noise. The metrics that matter in the search phase are unglamorous: retention, repeat usage, the percentage of trials that turn into habits. They move slowly, they’re small, and they tell the truth. \ What progress actually looks like Given all this noise, how do you know you’re getting anywhere? The honest answer is that progress in the search phase rarely looks like progress. It looks like a slow accumulation of small, qualitative shifts, most of which you’ll only recognize in hindsight. But there are things to watch for. The first is a change in the texture of conversations. Early on, you do all the talking. You explain, you demo, you handle objections, and the prospect mostly nods. Somewhere along the way, if things are working, the balance shifts. Prospects start finishing your sentences. They describe their problem using language eerily close to your positioning, sometimes before you’ve said it. They ask questions that presume they’re going to use the product (“how does this handle our SSO setup?”) rather than questions that audit whether they should (“so who else is using this?”). When the conversation starts pulling instead of needing to be pushed, something has changed. The second is the boomerang. Someone you talked to months ago, who went quiet, comes back. Or someone you’ve never met shows up having heard about you from a customer. Unprompted return and unprompted referral are among the strongest signals in early-stage life because nobody fakes them out of politeness. The third is repetition in the feedback. In the early days, every customer conversation surfaces different complaints and different feature requests, which is its own kind of signal: it means you haven’t found a coherent market yet because a coherent market wants coherent things. When the feedback starts converging, when the fifth and sixth and seventh customers all ask for the same thing and describe the same pain, you’ve located something real. Even if the convergent feedback is negative, it’s a gift, because at last the experiment is producing legible results. The fourth, specific to message/market fit, is theft. Customers start using your words. They describe your product to their colleagues using the framing you developed, sometimes verbatim. When your language escapes your own marketing and starts circulating on its own, the message has fit the market. Until then, every description a customer gives of your product is data: the words they choose, the comparisons they reach for, the parts they emphasize and the parts they leave out. Their language is usually better than yours. Steal it back. \ How to conduct the search without losing your mind There’s no formula for this, and you should be suspicious of anyone selling one. But there are postures that make the search faster and saner, and they’re worth naming. Write your hypotheses down. This sounds bureaucratic and is actually liberating. Before each iteration, state in plain language what you currently believe: who the customer is, what their problem is, why your product wins, and what words you’re betting will land. The point isn’t formality. The point is that an unwritten hypothesis mutates silently to fit whatever happened, and then you learn nothing. A written hypothesis can be wrong, and being wrong on paper is how you make progress. Six months in, the document also becomes a record of how far your thinking has traveled, which on the bad days is worth more than you’d expect. Change fewer things at once. You can’t run controlled experiments, but you can avoid maximally confounding yourself. If you rewrite the home page, reposition the product, change the pricing, and target a new segment all in the same month, whatever happens next will teach you nothing, because you won’t know what caused it. Sequence your changes when you can. Hold the product steady while you test a new message; hold the message steady while you test a new audience. Spend more time listening than building. This is the hardest discipline for technical founders, because building feels like progress and conversations feel like a distraction from progress. It’s exactly backwards. In the search phase, the scarce resource isn’t code, it’s information about why people do and don’t buy, and that information lives exclusively in the heads of your prospects and customers. Every week that passes without direct customer conversation is a week of flying blind. And when you do talk to customers, resist the urge to pitch. Ask about their problem, their current workaround, and the last time the pain flared up. The questions that begin “would you use a product that…” are nearly worthless because people are terrible at predicting their own behavior and excellent at being agreeable. Ask about what they’ve already done, not what they might do. Define, in advance, what would change your mind. Persistence is a founder virtue right up until it becomes denial, and the line between the two is invisible from the inside. The protection is to decide, while you’re still clear-headed, what evidence would convince you that a hypothesis is dead: if we can’t get ten paying customers from this segment in ninety days, the segment is wrong. You won’t always honor these tripwires, but having them at least forces a real conversation when you cross one. And calibrate your expectations about time. The honest range, for most companies, is one to two years of searching before product/market fit, sometimes longer, and the search for message/market fit continues in some form forever, because markets shift and language wears out. If you budget six months for this phase, you will spend month seven onward feeling like a failure for what is, in fact, the normal pace of the work. \ You are not executing yet. You are searching. Here, finally, is the thing I most wish someone had told me to tell first-time founders, because it changes how the whole frustrating period feels. Most of what you’ve absorbed about running a company, from business books, from press coverage, from watching big companies operate, is about execution: setting goals, hitting numbers, scaling what works. Execution assumes you know what works. In the pre-fit phase, you don’t, and pretending otherwise is the root error behind most early-stage flailing. Founders set growth targets before they know what they’re growing. They hire salespeople before they know what sells, marketers before they know what message lands. They scale the part of the company that’s supposed to amplify a signal before the signal exists. Steve Blank built an entire methodology around this distinction: a startup, in his definition, is not a small version of a big company but a temporary organization searching for a repeatable business model. The search and the execution are different jobs requiring different mindsets, and the failure to notice which job you’re doing is fatal more often than any competitor is. Once you accept that you’re searching, the frustration recontextualizes. The ambiguous signals, the discarded positioning, the pivot that erased three months of work: these aren’t evidence that you’re failing at execution. They’re the ordinary texture of search. A failed experiment that kills a bad hypothesis is progress, even though it feels like loss. The landing page that converted nobody told you something true. The segment that wouldn’t pay narrowed the map. Search progress is measured in eliminated possibilities and sharpened hypotheses, not in revenue, and if you measure this phase by execution metrics you will conclude, falsely and miserably, that nothing is happening. Something is happening. It’s just slow, and it’s supposed to be. Product/market fit and message/market fit are not milestones you schedule; they’re discoveries you earn, one ambiguous conversation at a time, by staying curious slightly longer than it’s comfortable to. The founders who get there are rarely the ones with the best initial idea. They’re the ones who treated their idea as a draft, their words as experiments, and their customers as the only authority that matters, and who kept running the loop after the romance of launch day wore off. The story you should tell yourself isn’t “build it, launch it, grow it.” It’s “guess, test, listen, revise.” Less cinematic, admittedly. But it has the advantage of being how it actually works. If this hit home, consider subscribing to Marketing for Grownups . It's a newsletter for operators who build categories and go to market without the hype. \