“Make Something People Want” ;)
Maybe your eyes glaze over when you hear someone say that, because it’s almost vacuously true.
But some of the best advice in life is simple, and some of the simplest advice is the hardest to follow. Beginning meditators are told to “just count their breath” — many spend years practicing. (If that surprises you, try doing it for 5 minutes without getting distracted.)
Likewise, many founders spend years learning how to “make things people want.” There’s a lot to unpack in those four words. Their concision makes them a useful short-hand: a principle you can use to guide a decision or meeting, an amulet you can rub when you’re wondering what to do. But concision is lossy. Abstraction sheds context and becomes its own jargon. It’s hard to know how to act on abstract, concise advice.
“Make something people want” is basically the startup version of “what would Jesus do?” In fact, it might be better phrased as a question, even at the price of concision, because that will set us on the trail of the answers, since a question wants to be answered, and answers must be sought.
Indeed, “make something people want” actually contains several questions: Which people? What something? How will you get it in their hands once you made it? And after you write down the answer to each of these sub questions, you might ask: how do I know that?
Having vague answers to these questions is often a sign that you are lost. Don’t worry — it’s OK to be lost! I call those vague answers “thinking in crayon.” They are directional, but the details have to be filled in. Getting too precise too soon is a failure mode. When you don’t have great answers to big questions, your real enemy is premature certainty, which is an easy refuge in the face of uncertainty’s discomfort.
Many people shrink from the idea of being lost and will do anything to avoid being perceived as lost by others. Silicon Valley puts a premium on confidence, and so does the business world more generally. Without certainty, what’s your story — what’s the adventure and dramatic arc of your life? But for pre-product-market fit founders, defaulting to confidence is a huge mistake that costs them years of their lives, and many great opportunities.
If you are already committed to a business plan and a product before you get the answers, then you are probably wasting your time, which is much worse than being lost, because in that case, you don’t even know you’re lost. If you’re lost, the first thing you should do is acknowledge that fact to yourself, act like it, and focus on finding answers to make yourself less lost.
The problem, of course, is that building something concrete can be fun and intrinsically motivating, while searching for answers to fundamental questions may feel uncomfortable and embarrassing, so it is easy to trick yourself into thinking you’re not lost by glossing over the hard work of seeking.
Another problem is that many cultures, including that of Silicon Valley, don’t reward lostness when they meet it at a dinner party. Investors and talent are attracted to vision, certainty, and momentum. But there is nothing worse than false certainty, or momentum going in the wrong direction.
So in order to “make something people want”, we really need to back up and start asking fundamental questions about what we’re doing. That is, if you agree with “build something people want,” then you should also concede that “it’s OK to be lost,” and therefore you must learn how to be lost in a promising way.
The main reason why many startups fail is because the founders do not take the time to be lost, but instead commit themselves too soon to a path that leads nowhere. If they catch a wave of tech hype, that may help them raise funds since they found PR-investor fit, but it often obscures the true path they must take, because they have made social commitments to a product and plan, and probably hired too many people to execute that vision, trapping them in a cage of their own promises and sunk costs.
If that is true, then the best thing we can do as a startup community and tech industry is find ways to navigate that dark and trackless forest, and to frame it in a way that helps people keep moving. When people say that someone else is lost, it’s often implied that they are clueless, maybe even too dumb to find clues. It’s also sometimes true that lostness, neediness and an excuse mindset are correlated. We don’t want lostness as an equilibrium or a pity party, we want the dynamic form that leads somewhere.
It’s possible to be smart and lost: To be lost in an intentional way, leaning into it, rather than using it as a signal for help, or as a cover for inefficacy. And the smarter we get about being lost, the greater the things we’ll probably find. Intelligence is defined by what you do, when you don’t know what to do. This used to be called searching.
Searching for PMF the Hard Way
Anyone who came of age after 2000 has to escape a language trap Google set for us. For most people, the word "search" means typing words in a text box, and getting results a second later.
While much knowledge is just a few keyboard strokes away, lots of things you want to know require more effort than that. They're not public, or they're not digitized, or they don't even exist yet -- the knowledge you seek can only emerge from you, a researcher and experimenter, acting on the world and gauging its response.
Historically, searching was a heroic and sometimes crazy act about which sagas were written, precisely because it was dangerous and rare (think: Jason and the Golden Fleece). Good old-fashioned searches are hard. As a result, few people embark on them, let alone find what they seek and make it home safe.
Steve Blank defined startups as “organizations in search of a business model” (PDF). While that is a true and valuable statement, unfortunately it's also a trap. Because you don't need a startup when you start out. Incorporating and raising money and building a team to go off and search for a business model is actually failure mode. (I think of this as a bad traction to obligation ratio. You want that ratio to be as high as possible to maintain your mental health as a founder. Going heavy on obligation without the traction is a trap.)
You want to do as much of your searching as possible pre-incorporation, without any legal entity or funds or team, because when you're searching, you need to travel light and allow yourself to be lost. Raising money and hiring people forces you to show a kind of confidence that is fatal to successful searching, because we all become the person we pretend to be.
It's harder to be lost when investors and employees look to you for clarity and leadership, and when you have put a stake in the ground about who you are and what you believe. So avoid that if possible!
The purpose of this post is to describe how someone can create the ideal conditions for search, and then motivate themselves until they find the first glimmer of product-market fit, with all that implies: who has a problem? how can it be solved? will anyone pay you to solve it? how can you build it?
This journey happens mostly through conversations and observation, you're traveling through minds and workplaces and scenes of life. You are visiting other peoples' worlds either by asking them questions, or observing them in action. (Jan Chipchase’s book “Hidden in Plain Sight” describes the process of observation well.)
Recently, a software engineering manager told me how a founder convinced him to join him in building an app and a business: the founder had had 400 coffee meetings over the previous year to find a group of people with an acute problem that his app would solve, and the budget to pay for it. That’s the search, and the effort that goes into the search. This founder didn’t have an idea, build the product, and conjure a market into existence. He had an idea, sought out the market, adapted his idea, and built the app.
Why Is Search So Hard?
To search well, you have to live a long time with uncertainty, while trying very hard to plow through unfamiliar territory. If you mistake this plowing as movement without purpose, the wind will go out of the sales (of your plow…? ;).
You have to let doubt ride in the sidecar to your hypothesis. Most humans are not pre-disposed to do that well or long. We are averse to doubt and more comfortable with certainty, which enables consistency and group belonging. Being lost feels dangerous.
The challenge then is to make out of doubt and search a meaningful process by which you can track your progress. You have to be a scientist whose life is itself the laboratory. What do you think is true and how can you test it quickly and cheaply?
First, write down what you think is true about a problem that you imagine exists; eg dog groomers need a new kind of AI to improve their customer experience. (Think: LLMs generating texts to respond to customer inquiries…)
Then write down the steps you could take to validate that hypothesis:
draw up a list of dog groomers near you
call or visit them in person
ask them about their customer journey
What are the stages a prospect moves through from the point of discovery (“Oh hey this dog groomer exists!”) to a sale (“I should pay them money because they're the best fit for me.”)?
(Creating ad hoc CRMs in a spreadsheet, tracking contacts and your interactions with them, is probably half the trick of orienting yourself while you are lost. With other major tools being journaling, and white boards with lines, nodes and exclamation points that you revise every so often.)
Each conversation with a dog groomer is a step on your journey. Each one puts you in contact with the kind of person you imagine has a problem, and allows you to get quick feedback on your mental model of how they work and what they need. Each moment of exposure and opportunity to learn is progress, and by structuring and recording your search journey, using whatever spreadsheet, journal, CRM or gold stars you have at your disposal, you can mark your distanced traveled.
This is important, because the uncertainty and indeterminate length of search make many people give up and go in search of easier dopamine hits (a quick news article, a flame war in your favorite messaging group, an Instagram account you have sworn off). Building a process that contains inherent rewards is how you incentivize yourself to push on.
This kind of dopamine engineering is a meta-skill that can support you doing many hard things (set goals, track progress, reward yourself for it, eliminate addictive distractions, raise your neurochemical baselines by getting healthy).
You can structure your search by breaking down the tasks to create the feeling of progress; even finding out that your hunch was wrong is progress, maybe the best kind. And you should know, there is no easier way. There's no shortcut. There's no syllabus out there. The only grade you'll get is in a few months or years, if you manage to get paying customers. You are inventing and revising your lesson plan one step at a time. And the only way you'll know what you learned is by taking careful notes.
The process of tracking your experiments, and creating a sense of progress, is analogous to (but more complicated than) tracking your hits and misses in a game of Battleship. Sure, there are more dimensions than just X and Y coordinates when you try to validate customers and problems, but without a record of what you tried, you will repeat your mistakes.
So … after taking detailed notes from your interviews and visits, you might notice patterns.
First, you may find that these businesses aren't ready for AI yet. They're still getting used to the idea of data, and don't really know what to do with software. Some of them probably operate in the dark; that is, they don't actually have visibility into their crucial processes.
Maybe they can't tell you how many website visitors convert to a chat (if they even chat with website visitors), or how many chats convert to a paid session at the groomers. Maybe they can't say how many grooming sessions each member of their staff completes per day, whether there are differences among the personnel, and what explains those differences. That is, they barely know how their business is doing, let alone which levers to grab to improve it.
Notice, here, that you started out thinking you were in search of answers, but the answers you got actually taught you how to ask a better question. Deep down, you were actually in search of the right question all along, and the weird thing is, the wrong question can lead you to the right one. So your progress “forward” actually leads you “backward” a step toward the right question, which unlocks everything.
That is, you have this stacked search, the explicit one and the implicit one. And if you embark on the explicit search right, it will lead you to question your assumptions and relaunch your search with a better frame. (You may even find that dog groomers do not have a problem that your skills are well suited to solve, in which case, the answers you find will push you even further back to exploring another type of business altogether.)
All this exists in a search hierarchy of Business sector/User-with-problem/Potential-solution, where we tend to ask questions about solutions before we truly understand the problems, and then ask about problems before we understand the whole user, their context, and what they really value or must overcome. And we must understand all of that to know if we're chasing the right opportunity.
(An analogy to this hierarchy: Imagine you’re a bed repairman searching for a bed to fix in a blacked-out room: you don’t know if the bed exists, or if it’s broken, or where the walls are, or if the room is even big enough for a bed to fit inside (hello, market size)! So you’re modifying the hypotheses your search is based on in real-time as you grope in the dark.)
The word for this in English, not widely used, is “zetetic”, which derives from the Greek verb zetein (to seek) and means “proceeding by inquiry or investigation.” The path to product-market fit, which goes first through customer validation, is zetetic. The word “seeker” is spiritually loaded. The word “scientist” is too academic and overemphasizes knowledge. Here we are talking about the art of not-knowing. Zetetic is better than “skeptic,” because skeptics do not have an inherent forward motion, and better than “agnostic,” because agnostics have sworn off knowledge and belief. Zetetics are people who move forward skeptically through uncertainty toward belief: They have a rough goal and a process for gathering and gauging evidence, and they must decide without perfect knowledge. They are in a perpetual cycle of imperfect knowledge and necessary decisions as they attempt to true their course.
Going through the search motion of problem/value hypothesis, target discovery, tons of conversations, learning from your notes, backing up, more conversations, backing up again, more conversations, and maybe finally an MVP to test real behavior, is something you'll need to do over and over. That's the dotted line crossing the parchment of the map, which is covered with an imagined bestiary because no one's ever been there before. One great method to use is what Gagan Biyani calls the “Minimum Viable Testing Framework”. (Please stop reading this post and read his instead — it’s excellent!)
One mistake that is much costlier than asking about a potential solution before you really grasp the problem is to BUILD your castle-in-the-sky solution before you really find out what people need. Engineers do the build-first ask-questions-later thing all the time, because they love to build and the dopamine trough of uncertainty and lengthy interaction with strangers makes them sad and anxious. This is a good way to waste years of your life on something no one needs.
There are two parts in an engine that can get to product market fit: 1) the ability to understand and learn from users with a problem and budget; 2) the ability to build something that helps them. Sometimes, founding teams separate these parts into two roles. Other times, a solo founder can both explore and make.
(Once you have validated your sector, the problem, the user's budget and others things, you may find that indeed the product is the conversation; i.e. adding a quick new feature is a good way to find out how and whether people will use it, since users are bad at predicting their own future behavior.)
To take care of part no. 1, you can either embody the user yourself, or you can embrace the intense and gregarious learning process of reaching out to inhabit other peoples' lives. But only if you are lucky enough to find some people who will give you the time of day.
Your method for finding and extracting information from the supposed problem-havers will vary by industry and problem. Maybe the problem-havers are people in your social network, current or former colleagues, alumni of institutions you belong to, fellow members of club or community, family or friends. Sometimes you can network into communities of people with a problem you think you can solve.
Other times you have to pay people to share what they know. Userinterviews.com will help you find certain types of users. For folks with a bigger budget, expert networks like GLG will take your money and conduct hundreds of interviews for you. (While you should get feedback from the market however you can, if you are paying people for it, that is a sign that you need to do more work immersing yourself in their world, until your social network and experience provides much of the feedback you need.) Still other times, with the goal of immersion in a new industry, the future founder should just get a job in their target industry to learn and network on the inside.
A parting thought: Even successful founders and profitable companies have to allow themselves to be lost again when they go in search of a new product or business line. Many of them do that poorly because they have an operating culture that cannot tolerate uncertainty or failure, or because they have become blinded by the success of their previous products and confused luck with skill.
Admitting you’re lost is the first step you can take to touch reality.