Founder of Qulture.Rocks (YC W18) and on a mission to help teams unlock their potential!
Many people talk about product-market fit, and how important it is in the early days of a startup. Actually, most people agree that product-market fit is fundamental to a startup’s trajectory.
But it’s amazing how little agreement there is around what “product-market fit” actually means. Most people can more or less agree on how it feels to have product-market fit, but not on what product-market fit really is.
In this article, we’re going to discuss product-market fit: what it means, why it’s important, and most importantly, what’s the right path to take in order to attain it.
The expression “product-market fit” was, supposedly , coined by Andy Rachleff, founder of Betterment, but made famous by Marc Andreessen, founder of Netscape, Opsware, and Andreessen Horowitz, in his The Only Thing that Matters post, where Andreessen discusses why he thinks the quality of the market  a startup is attacking is the most important variable in its success equation (the other two legs of the tripod are team and product).
Andreessen says that Product/market fit means being in a good market with a product that can satisfy that market. His definition is kind of recursive because he goes on to define a good market as a market with lots of real potential customers — the market pulls product out of the startup.
Andreessen goes on to talk more about what he means by “pull” when he explains what product-market fit and its absence feel like to a company:
“You can always feel when product/market fit isn’t happening. The customers aren’t quite getting value out of the product, word of mouth isn’t spreading, usage isn’t growing that fast, press reviews are kind of “blah”, the sales cycle takes too long, and lots of deals never close.
And you can always feel product/market fit when it’s happening. The customers are buying the product just as fast as you can make it — or usage is growing just as fast as you can add more servers. Money from customers is piling up in your company checking account. You’re hiring sales and customer support staff as fast as you can. Reporters are calling because they’ve heard about your hot new thing and they want to talk to you about it. You start getting entrepreneur of the year awards from Harvard Business School. Investment bankers are staking out your house. You could eat free for a year at Buck’s.”
Andreessen says that when a startup is tackling a great market and manages to make a product that solves its need, the market will almost effortlessly pull the product from the startup; the product will almost sell itself, and growth will be so intense that it breaks the organization over and over while it grows to capture the opportunity. It’s almost as if “great” is the product of “big” and “where the product is pulled.”
Two interesting things about Andreessen’s view of the whole “product-market fit” thing.
First, it’s interesting that he says it’s not enough to have a good solution for a big, rich market: it’s also crucial for the product to be pulled. That means, among other things, that users are prone to getting the problem solved now, but also that they actually do it. There’s a big problem with that, which we’ll discuss later: since "pull" is a lagging indicator, it’s very hard to optimize for it. The company may identify a big market, build something that solves its need, but end up not feeling the pull.
Second, it’s interesting that Andreessen doesn’t mention what happens after users have purchased the product: he doesn’t talk about churn, retention, expansion, or engagement, for that matter. So by Marc Andreessen’s definition, product-market fit is when a startup has built a product that sells a lot and that has the potential to sell a lot more. It’s almost a go-to-market thing.
A better definition, that would better convey what Andreessen and Rachleff were thinking, is product/great market fit, where a great market is, as we’ve seen, defined by its size, by the amount of pain it has, and by the willingness of the market to buy a solution now for that pain.
Anyway, putting two and two together, we get a more complete Rachleff/Andreessen definition:
Product/market fit is when a startup has built a product that can satisfy a market that’s big (has many potential customers) and that easily and intensely buys/uses that product.
What we can conclude from the Rachleff/Andreessen is that the term “product-market fit” is not a very good choice to describe what they meant in the first place. A better term must convey the idea of market “goodness,” since it’s not enough to have a product that fits any market. The market has to be good. A better definition, that would better convey what Andreessen and Rachleff were thinking, is product/great market fit, where a great market is, as we’ve seen, defined by its size, by the amount of pain it has, and by the willingness of the market to buy a solution now for that pain.
Nowadays (the article was originally written in 2007), the term product-market fit has been used in the startup world with many similar but not quite equal meanings.
One common use of the product-market fit concept “disregards” the greatness of the market inherent to Andreessen’s original use. By the way, in our experience, this is the most common way people use the term “product-market fit” nowadays, and frankly, this meaning does derive more naturally from the term. Anyway, under this definition, product-market fit, or PMF, for short, is when you’ve built something that some set of users (a market), regardless of how small it is, love.
Paul Graham, of Y Combinator fame, calls this “making something people want.” In his “How to Get Startup Ideas” essay, which is amazing, he explains why startups should find small niches within big markets and find love from users in these niches before venturing out to a broader market:
When a startup launches, there have to be at least some users who really need what they’re making — not just people who could see themselves using it one day, but who want it urgently. Usually, this initial group of users is small, for the simple reason that if there were something that large numbers of people urgently needed and that could be built with the amount of effort a startup usually puts into a version one, it would probably already exist. Which means you have to compromise on one dimension: you can either build something a large number of people want a small amount, or something a small number of people want a large amount. Choose the latter. Not all ideas of that type are good startup ideas, but nearly all good startup ideas are of that type.
Imagine a graph whose x axis represents all the people who might want what you’re making and whose y axis represents how much they want it. If you invert the scale on the y axis, you can envision companies as holes. Google is an immense crater: hundreds of millions of people use it, and they need it a lot. A startup just starting out can’t expect to excavate that much volume. So you have two choices about the shape of hole you start with. You can either dig a hole that’s broad but shallow, or one that’s narrow and deep, like a well.
Graham goes on to explain why he thinks “market” is not as big of a factor to optimize for . First, because markets evolve and usually have many adjacencies, and some of the best startups have tackled markets that looked small initially but that went on to become huge, either on their own or by merging with nearby adjacent markets. Second, because entrepreneurs most probably won’t be able to find a clearly amazing market that’s still untapped — chances are somebody will have taken it on already. Third, that the true bottleneck is getting users to love something. That if you can do that, you can figure the market problem later.
Nearly all good startup ideas are of the second type . Microsoft was a well when they made Altair Basic. There were only a couple thousand Altair owners, but without this software they were programming in machine language. Thirty years later Facebook had the same shape. Their first site was exclusively for Harvard students, of which there are only a few thousand, but those few thousand users wanted it a lot .
When you have an idea for a startup, ask yourself: who wants this right now? Who wants this so much that they’ll use it even when it’s a crappy version one made by a two-person startup they’ve never heard of? If you can’t answer that, the idea is probably bad.
You don’t need the narrowness of the well per se. It’s depth you need; you get narrowness as a byproduct of optimizing for depth (and speed). But you almost always do get it. In practice the link between depth and narrowness is so strong that it’s a good sign when you know that an idea will appeal strongly to a specific group or type of user.
It seems that PG and Rachleff/Andreessen are saying different things. And they might because they might have different motives.
By this definition, PMF means having a product that fulfills the need of a market, regardless of the quality of that market. We think this is a good term to use if you keep in mind that it doesn’t actually convey what Rachleff and Andreessen meant in the first place.
PG clearly writes to help entrepreneurs like the ones who go through Y Combinator. He’s giving out advice that compensates for some common founder biases or mistakes, like not spending enough time listening to users/customers, and/or spending too much time thinking about grand strategies . Andreessen and Rachleff, on the other hand, sound like they are just musing on what makes, in hindsight, a great valuable company (a unicorn, in today’s parlance).
If we agree to use “product-market fit” meaning “product/great-market fit,” it’s easy to see why it’s an important thing to achieve: it is the recipe for building a large venture-backed company.
Now for the hard part.
If product/great market fit is a proxy for becoming a unicorn, how can founders get to it? What’s the recipe for finding product/great market fit?
The answer is simple. First, find a great market. Second, find product/market fit within that market.
Simple, but not easy. Finding product/great market fit is an incredibly difficult challenge, and maybe it’s so hard that PG purposely asked us to ignore the “market” thing.
In order to find product/great market fit, you’d have to start by looking for a great market. And to look for a great market (we defined “great” before) you need to first look for a big market. That would be “a large group of users with an unmet need.”
The second step would be to build a product that satisfies that need, or “to make something people want”.
After you’ve found a big market, you have to figure out if it’s a great one. The only way to truly know if you are in a great market is to actually be inside that great market, experiencing it. And that’s only possible ex-post, or after you’ve built the product and sold it successfully.
The big problem is that you can’t really figure out if a big market is willing and able to solve its pain now before you have a product that’s being sold like hell.
You are stuck in a circular reference.
You must find out if a big market is a great market in order to start building your product, but you can only find out if a market is great after you have a product that sells a lot in that market. So entrepreneurs have to rely on educated guesses and hunches and take on the risk of not finding a great market. In a nutshell, it’s not a bad way to describe what the startup journey is all about.
It’s important to know that you will probably not know for sure you’ve found a great market until much later on the process after you’ve built a product. And if you don’t find it you’ll never know who was to blame, the product or the market, since they’re both sides of the same coin.
There are some tools that can help entrepreneurs make better decisions.
One way consumer or SMB-focused entrepreneurs have used to make progress on that front is through “false” search ads and landing pages, that serve as a first test to gauge if people are even willing to sign up for a (nonexisting) demo or join a waitlist. A high number of users sharing their emails hardly proves if a market is great, but may be an easy way to cut losses quickly if users don’t even do it.
Another technique entrepreneurs in the enterprise space have used to figure out if a market isn’t that great is to have customers sign letters of intent (LOEs) stating their intention to buy a product from the startup for an approximate amount of money if it succeeds in solving some pain point.
The second step would be to build a product that satisfies that need, or “to make something people want”. (By the way, a great article that describes the “make something people want” journey is Rahul Vohra’s How Superhuman Built an Engine to Find Product/Market Fit).
The process of making something people want is also quite simple:
Step 1 is pretty easy. If you’re starting a startup, chances are you can build stuff, or have someone on your team that can build stuff.
Step 2 is also quite straightforward. You must find users to use your product. This varies from consumer/B2C to SMB to enterprise, and there’s a lot of stuff on the web about it. Consumer startups may “launch” on Hacker News or do a PR blitz to get a bucket load of first users in.
Step 3 is a bit trickier. You must find out a) if these users really get value from your product and b) if they are willing to pay you. Retention may be a great proxy for value. Some, like Vohra and Sean Ellis, who inspired Superhuman's approach, propose you ask users how frustrated they would be if they couldn't use your product anymore. But to find out if your users will pay for your solution, on the other hand, there's no alternative to asking them to part ways with their cash .
If you don’t have retention, you have to learn why, so that you can go back to Step 1 and try something different. But humans don’t always communicate clearly about why they do stuff. You can’t know for sure if the reason they’re giving you is the real reason they aren’t using your product. (A great way to make this process easier is to make something you want. To solve a problem/pain you have.)
There’s lots of art and little science in making something people want.
If you’ve made something people want and started out with a big market, that can possibly be great, then the only step left in order to reach Rachleff/Andreesen nirvana is to validate if it’s a great market indeed.
Startups that have clearly found that nirvana will know it instinctively: as Emmett Shear, founder and CEO of Reddit, puts it:
“What is “product/market fit”? I’m not sure I can give you a definition. But maybe I can share what the subjective difference is in how it feels when you have it and when you don’t. Founding a startup is deciding to take on the burden of Sisyphus: pushing a boulder up a hill. If you stop pushing, the boulder doesn’t move. Every inch you push the boulder requires full effort. If you get distracted or stop working, the boulder might actually roll back down. You’re sweating and shoving and progress comes slowly and incrementally…. But there is this moment that comes, eventually, where you crest the top of the hill. And suddenly you’re pushing the boulder and it’s moving so easily! You’re pushing it on the summit, and the hill doesn’t fight you. It takes only a little effort to make progress. You have reached the promised land of product/market fit. And now the boulder is starting to roll down the hill, you actually don’t need to do any work to make progress. Customers are coming to you, you have more demand than you need. And the boulder starts to accelerate.”
According to Emmett (and to Andreessen), growing will feel like pushing a boulder up the hill. Solving it, though, is very hard. You may feel you’re pushing a boulder up a hill because of two main reasons. Let’s call them category 1 and category 2:
Category 1: The market is indeed great, but your growth engine is broken, orCategory 2: The market wasn’t great to begin with
The first category (growth) is easier to solve. You can tweak your growth loops, try new acquisition channels, etc. The second category is harder to solve (you can’t change your users, at least don’t count on it): the only solution is a pivot, which is always very risky since you’re almost starting your startup from scratch.
The beauty is you can’t ever know for sure if you’re dealing with a Category 1 or 2 problem. You can always grind a little more, try some different acquisition tactic or an adjacent market.
But you will never know for sure.
 According to Google Trends, the first searches for “product market fit” happened in 2005, as you can see below:
From that, we can derive that Andreessen’s post wasn’t the first time the expression was used. We did find one other hit from 2005 that seems promising:
 We love the jobs-to-be-done definition of a market, that says that a market is the sum of all customers (people or organizations) that have a need and that are open to using a solution to fulfill that need. And a great market is a) huge (size), b) has people/organizations with intense needs (pain), that are c) very open to using something to solve that need.
 To be clear, he doesn’t say markets are not important. He just says that if you’re starting a startup, the most important bottleneck to solve is getting even a small set of users to love what you’re making. If you can do that, you will most probably be fine.
 Second type = build products that are deeply loved by a small set of users.
 I don’t really agree with PG on the use of Facebook as an example. I think it’s quite a stretch to single out “Harvard students” from “college students” as separate markets. I think college students are pretty much like each other, and a pretty large market that could have made for a big company. I think Facebook’s example is a great example of the right tactics of building network-effect products, like social networks and marketplaces. By almost artificially fencing your audience, as Zuckerberg did with Harvard, it’s easier to build a critical mass for that audience.
 In some cases, especially around B2C, services are free for users, and monetized alternatively (think Facebook and ads). So you won’t be able to find out if you can make money from the business until later on. For those businesses, retention metrics are enough.