I’ve been spending a lot of time trying to wrap my head around Product/Market Fit (PMF) recently, in order to understand what the goal is, how to get there, and how to apply it to the projects that I’m working on everyday. It has been a great process that has ended up sparking a ton of insight and creativity in my work.
I’ve become so immersed in thinking about PMF that I wanted to write down my thinking to crystallize the lessons from everything I’ve read to date… below is what came out. I hope it helps you think about building better products, like it has me.
What is Product/Market Fit?
The quintessential quote defining PMF is from Marc Andreessen, in part because he’s the one who coined the term back in 2007, but also because, well… it’s Marc Andreessen:
“Product/market fit means being in a good market with a product that can satisfy that market.”
In this original article about PMF, Andreessen discusses what’s more important when looking at a startup’s potential: the product, the team or the market. In his view, the market opportunity is first and foremost, and the discussion results in the above definition of PMF as the goal to strive towards: i.e.: if the market is the most important element, than having a product that truly satisfies the market’s needs becomes the objective.
I prefer to look at things from a Human Centred Design perspective, which I think really helps to unpack what Product/Market Fit actually effectively means. To me, it’s the intersection between:
- A problem that a sizeable group of people really need solved = i.e.: Desirability
- A product that can actually be built well to fully solve that problem = i.e.: Feasibility
- A business model that can be executed to be profitable at some point in time = i.e.: Viability
In my mind, If you don’t have all three of these elements, you haven’t reached PMF.
I’ve been finding it useful to think about PMF with this model in mind, because it gives me something more tangible to work towards. I can therefore measure the progress of my various projects to where they are in achieving a relative ‘check mark’ for each element, in order to help guide future testing and development.
Always in my mind now is (1) what do I have, (2) what am I missing, and (3) what do I need to do now to get me closer?
How can you judge if you have reached PMF?
There seems to be a lot of posts out there that try to give you “the number” you need to hit to say you have PMF… a specific ARR in SaaS businesses, for instance. You also read about how most of that is BS, so I’ve been ignoring specifics. The pervasive, but kind of not helpful, view point is that true PMF is so very powerful, that you just know it when you see it
From reading way too many startup stories, I can imagine that… but this vague notion is just not implementable: if you know PMF when you see PMF, great; but how do you know if you’re on the right path? So I put together a list of metrics that help you understand your progress and determine where your gaps are:
- Net Promoter Score — do a significant proportion of your users love your product so much they are highly likely to recommend it to a friend? This is about judging how much people love your product. The more they love it, the more engaged they’ll be, the more they’ll pay you and the more they’ll help you get new users.
- Retained engagement — does a significant portion of new users start using your product, and then literally never stop (i.e.: the retention curve asymptotes to a solid % of users)? alexschultz from the growth team at Facebook discusses this engagement curve during his lecture at YC’s startup school. Worth a watch.
- % Users Driving Revenue — are a significant portion of users converting to providing you with revenue? (i.e.: by paying you directly or indirectly — depending on your business model) This is a judge of just how big the problem you’re solving is, and how good your product is at solving it. If it’s not a big problem, people won’t pay. If you don’t solve a big problem well enough, people also won’t pay (although, “well enough” is a relative term, and doesn’t mean a fancy, polished product. It is often highlighted that people are surprised at how raw their product is when they hit PMF. “Well enough” is about how well it solves the problem, not how good it looks or how many added features it has. It’s also relative to how big of a problem it is, and how manageable or horrible the alternatives are).
- Organic Growth % — is significant growth happening naturally? Are users so happy with your product that they’re telling everyone who’ll listen about it? Are potential new users so in need of a solution that they’re searching you out in droves? Is the promise of your value proposition so good that people convert with no effort? This is the number that apparently just takes off on its own, once you’ve truly nailed PMF.
That said, the above are generic metrics that may not be fully applicable to your situation. Think through what metrics truly matter for how you can judge Desirability, Feasibility and Viability in your product, and go with that.
One final thought on metrics: keep in mind that Product/Market Fit isn’t about being able to say “things are going pretty decent, looks like we hit PMF.” It’s actually about “holy shit, things just took off”. If you’re not blown away by your performance, you’re not there yet.
“Founders tend of underestimate how well things work when they do start working.” rob go, NextView Ventures
An aside on more complicated models:
When I think about 2-sided businesses, like advertising models and marketplaces, it strikes me that these companies need to reach Product/Market Fit twice. For instance, in advertising models, you have (1) PMF for the users, and then (2) PMF for the advertisers… The process becomes reaching PMF for the users first, and then (once you have a solid base of traffic and interaction) you start to work on PMF for the advertisers. Ex: Facebook has both, Snap is working on #2.
This is similar for 2-sided marketplaces, only in those cases it’s chicken/egg: do you start with demand-side or supply-side? An interesting example of this in action is The Muse, which is a 2-sided marketplace for (1) people looking to advance their careers, and (2) companies looking to hire smart candidates. They had an great strategy to solve the chicken/egg problem. They created PMF for the demand-side (the workforce) for an entirely different product first, in order to build an audience. They did this with an adjacent value-proposition of content about career advice. Then, with a potential demand-side audience in place, they added in company profiles and job postings to build the actual product and business they were envisioning, and quickly reached PMF for job postings for both demand-side and supply-side simultaneously. Pretty smart.
What does the macro process look like in theory?
While it’ll never work out like this in the real world, I see the intended journey to Product/Market Fit (and then scale) as a 6-step process:
0 — Starting Point
You need some nugget, some pain point, some passion or hunch to spring off from. The early stages, starting here, are about understanding and validating significant Desirability, and therefore the focus is on understanding a big problem people have, not on coming up with novel ideas. If you’re stuck here, and don’t know where to even begin, be patient and open your mind to problems you experience everyday that annoy the hell out of you.
Paul Graham has great advice for this stage. He says to “live in the future, then build what’s missing.” By this, he means try every new product you can get your hands, read everything you can about trends and even read science fiction. Then, think about what hasn’t been solved yet that really should, and do that.
1 — Problem Insight
This is where you actually begin to work. In this stage, you basically perform a lean and mean and fast-as-hell research project. Talk to as many relevant people as possible, as quickly as possible, in person as much as possible to understand if this problem exists, if it is big enough for people to pay for it, if it is so big and juicy that some would actually pay you right now to solve it, and who feels this problem the most. There’s tons written about uncovering the “Job To Be Done” (my post, for instance), or the “User Goals” (About Face is the quintessential textbook on goal-oriented UX design). Work to understand these methodologies and do your homework.
Remember, this step is about uncovering a deep problem. Get truly insightful before you move forward.
2 — Problem/Solution Validation
Here is where you begin to think about Feasibility. In this stage, you want to have a hypothesis about what the output of a solution looks like, and test whether it actually solves the problem. This is not about testing the technical aspects of the process you believe will create that output, but whether the output itself solves the problem. Answer this question for yourself: can this problem feasibly be solved to a high enough degree for people to throw money at you for the solution output? The process to get there uses a Minimum Viable Product (MVP).
“An MVP is not a cheaper product, it’s about smart learning” steve blank
Steve Blank, tells a great story about an agriculture data company who were trying to prove that if farmers had better arial data about their fields’ health and moisture levels (the solution output), they would be able to solve the problem of wasting resources by watering their entire fields equally (the problem). The product they envisioned used drones, specialized cameras and complicated data-analysis software to provide the data output, and therefore thought the best MVP was a hacked version of that product, which would take a few months (and a bunch of cash) to build. Instead of this, Steve had them fly over a few fields in a helicopter with a camera, manually process the data and see if the farmers would buy the output and put its insights to good use. They were able to execute that over a weekend.
That’s what an MVP should be… a test of “if I provide you with this solution, does it solve this problem that I know you have?” It’s about the solution output, not the specific process that produced it. The challenge is to think creatively about what’s the best MVP for your product that tells you if your hypothesized solution output works, as fast and as cheap as possible.
3 — Product Iterations
Here is when you’re finally in a place to build and launch an actual product. Alpha version, private beta… public beta, etc.. Whatever the process you want to take to put your early product versions into market, here’s when you do it. In this stage, you continue to refine Desirability, broaden the work on Feasibility to executing an actual product that can create the solution output discussed above, and start to focus on understanding whether there is a Viable business model in here somewhere.
The key here is having an incredibly fast learning cycle as you work your way from “definitely not having PMF”, to an awesome and sustained “WOOOOOOO!!!!!!!!” that comes when you reach PMF (more on the learning cycle below). Start with the absolutely must-have core user experience, and build / tweak from there. It’s about learning what’s going to help you nail PMF, so a combination of analysis of your data (user behaviours), and ongoing qualitative-based user interviews is what gives you the insights required to get there.
4 — Product/Market Fit
Pretty straight forward here, because we already discussed it, but this is when everything just clicks, and it’s off to the races. It is the moment in time where you shift from a sole focus on creating the right product, to continuing to do that while actually building the business (aggressively so).
5 — Scale
Once you hit PMF, you need to triple down on scaling the product. You need to always be improving on the product offering, while optimizing the funnel, building out sales and marketing teams and scaling the organization to support.
A massive warning here is that you can lose PMF at any given time, due to competitive pressures, a screw-up in your product roadmap and even broader changes in the marketplace. So stay ahead of things and be rigorous, insightful and methodical about product development, and keep an eye on early-warning signs: NPS, engagement, % revenue and growth.
Again, in practice, the process is never this perfect and straight forward (more on that later), but I find it useful to have a simplified process map in my head, in order to understand where I’m going and how to theoretically get there.
The micro process: iterative learning
Throughout the entire macro process of working towards Product/Market Fit, there’s the micro process of iterative learning. This is typically described (for example, in The Lean Startup) as the “build, measure, learn” feedback loop, of which you want to run through as many times as possible, as fast as possible, in order to iterate your way to success.
That said, I read an interesting post recently by Tristan Kromer that challenged this, and said the process should be flipped to “learn, measure, build,” because “if your goal is validated learning, the first thing you need to do is decide what to learn.” So, his process becomes:
- Establish a hypothesis
- Determine a quantitative or qualitative method to evaluate that hypothesis
- Build an experiment to test that hypothesis
I’ve sort of combined the two, because I think he’s putting to much into his step #3. For me, the process that you should be running through over and over again, as fast as your team can handle and be effective, should be:
- Hypothesize about something you believe to be true
- Design an experiment for how you’ll test the hypothesis to be true
- Build that experiment and put it into market
- Measure the results of that experiment
- Learn what that experiment taught you
- Repeat the process by feeding those learnings back into a refined hypothesis, moving the testing along through the macro process as appropriate
You need to run through cycles of this process as fast as possible to ensure you’re learning and progressing without wasting precious runway. As such, they key is to balance putting your head down and methodically executing cycles, while keeping your head up to see the big picture so you can identify if you need to tweak or pivot. This is not an easy thing to do, but it’s so very important.
How the process tends to play out in reality
As briefly mentioned, the perfectly straight-forward-seeming process I described above almost never plays out that way in the real world. As most products and businesses search for Product/Market Fit, they seem to get on a crazy rollercoaster of ups, downs, loops and pivots.
rob go, of NextView Ventures, had an amazing post a few weeks back about just this, called “The Shape Of Traction” (as well as a follow-up podcast that went even deeper). His basic thesis is that while it’s obvious when you definitely have or definitely don’t have PMF… what’s difficult is the in between, where “if you squint” you could convince yourself you might be getting close to PMF, but you may be very far away.
What he encourages teams to do is to be radical in their testing to find PMF. Try drastically different ways to solve the problem, or completely different personas to target, or even different problems to solve with a particular technology insight. He thinks the time and resources for searching around for traction are best spent with wildly different hypotheses, vs executing the learning process to fine-tune the same core hypothesis. The later approach often ends up burning through your runway, with you realizing too late that you’re not going to find PMF. Testing radically different things gives you a much better sense of what is ‘meh’ and what is breakthrough.
There are tons of great examples of slight and immense pivots, here are just two:
- Mattermark — wanted to be a better technology content business, but stumbled onto selling the business intelligence data that it was collecting in order to automate mundane tech news.
(link to a podcast discussing this) - Goat — the team began by working on a product that would bring like-minded people together for dinner, but after a lot of time they couldn’t gain traction. They threw away their product and executed a massive pivot into a mobile marketplace for collectible sneakers.
(link to a post discussing this)
This is where you need an analytical mindset and to not fall in love with your ideas, but instead fall in love with the problem. If the problem is juicy enough, you’ll find a way to solve it… it’s just really not likely to be solved in the first way you try.
The importance of thinking BIG PICTURE
I’ve mentioned it in this post, and I’ve come across it time and time again: when you reach Product/Market Fit, you just know, because things just take off. If things haven’t taken off, you’re just not there yet — for instance, in the Rob Go article I mentioned above, he suggests that people fool themselves into believing they’ve reached PMF, because they’ve sliced the target user definition too fine by only focusing on the user groups that were responding positively to early versions of the product. Instead, teams should be focusing on how to make the product better, in order to address a larger market and work towards true Product/Market Fit (remember: market is the most important factor).
That said, I can understand why people can only judge the point in time when they reach PMF in retrospect, which Ben Horowitz (bhorowitz) seems to suggest. You can see this come to life in the fake-sample image of a product’s sales below, which at some point reached PMF and hit exponential growth. If you zoom in the point in the curve where things started to turn, it’s not very obvious that something remarkable is starting to happen, it just looks like decent, but normal results.
You can only really judge in retrospect.
Because of this, I think it’s key to think about the bigger picture. Zoom out and see what things look like. Take into account growth over longer time periods to see what’s going on. Use all of the metrics I mentioned above to see if everything is clicking, or if it’s just some elements that are doing well. If you’re looking at revenue alone (as per the above), it can be tough to tell, so view things holistically.
Remember, Product/Market Fit is about reaching a point in your development where you’re properly solving a real problem that a large market of people need solved RIGHT NOW, in a way that you can extract value to grow a true business. If you can’t honestly prove those three elements, then you aren’t there yet… you might be close, but you also might be far away and need to pivot.
Be insightful, believe in yourself, but be ruthlessly analytical.
So there you have it, how I now view what Product/Market Fit really is, the process to get there, and the nuance about finding it. As I said off the top, it’s been a great process to research and learn more about a concept I previously understood, but only at a high level. I hope that me writing all of this out helped you, even a fraction as much as it has helped me.
Please let me know what you think. Do you agree with my perspectives on PMF? Do have interesting thoughts, experiences or articles that can add context to the above? Would love to hear from you.
Thanks for reading!
B
PS: What I’m going to do now is to respond to this post with a bunch of the key articles that I read over the past few weeks about PMF. Not sure if that’s going to be helpful, or super weird, but I figured I’d try it out this time to see. Let me know what you think.
