Product market fit is a popular term. I believe it's misunderstood. Businesses or products don't attain it forever at any given point. You can just as easily lose product-market fit as you gained it. This post dives deeper into the concept and how you can measure product-market fit.
  • Published date
    July 30, 2022

Product market fit isn't a destination

Rand Fishkin, the founder of Moz and Sparktoro, declared Product-Market fit as a broken concept and a Customer Adoption Spectrum as a better alternative. I've been following Rand on Twitter and came across his opinions on this topic and was shocked. Here's a veteran poo-pooing the concept that startup founders live and die by. 

Naturally, I had to dig in and find out what he had to say.

In his article in late 2020, Rand says:

The startup ecosystem loves to hype the concept of product-market fit. Does your company have it? How far away from it are you? If you have it, how are you scaling? If you don't have it, why would you ever try to do marketing? All of these questions assume a binary truth, universally applicable to all companies and the products they make. In this mythology, it's either: Yes, you have product-market fit. Or, no, you do not have product-market fit.

He even took direct aim at Lenny's pieces on the same topic in his piece and went as far as saying this.

Lenny Rachitsky's recent newsletter "What it feels like when you've found product-market fit", reinforced this mythology. Reading through the many impressive answers from a variety of accomplished founders and product-builders, I struggled with how Lenny didn't reach the obvious conclusion: that there's no such thing as P/M Fit.

Ouch. Shots fired!

So is Rand right? Is product-market fit dead as a concept? Are all of the intelligent technologists who pioneered the concept of product-market fit missing the forest for the trees? 

The truth emerges when you move away from the click-baity absolutist nature of modern content marketing and think deeply about why any of these frameworks exist in the first place. In this article, I'll lay out what I believe product-market is all about and why it's misunderstood despite all the excellent writing on the topic.

What is product-market fit, and why should you care?

Some basic googling indicates that the term was first coined by Don Valentine himself - the founder of Sequoia Capital, the godfather of the startup ecosystem in Silicon Valley. It was later popularized by Andy Rachleff, Steve Blank, and Marc Andreesen. The concept has existed in the startup vocabulary for a while. There have been various ways its been described. Lenny's article did a great job of outlining the multiple definitions in the modern era, so check it out. 

In summary, the concept can be understood in the following way. A startup has product-market fit when there is an overwhelmingly strong pull from the market for the product that the startup is selling. VCs need something like this to assess whether a startup will make it before they decide to employ capital in the pursuit of return. Startups readily adopted this concept to ensure they speak the same language as VCs when pursuing capital.

However, a challenge remains - this definition isn't easily measurable. How do you measure overwhelmingly strong pull?

Enter Sean Ellis and Rahul Vohra. Sean Ellis, the founder of Qualaroo, an early marketer at Logmein, Dropbox, and Lookout, and the coiner of the term Growth Hacking, came up with a way to measure product-market fit. Rahul Vohra, of Rapportive and Superhuman fame, popularized this measurement strategy when he wrote about how he used it as he developed and took Superhuman to market. I remember reading Rahul's post multiple times when it first appeared in First Round Review. Rahul and his team methodically measured product-market fit and changed their product and go-to-market based on the lessons. It's still such a good read.

So how did Sean and Rahul measure product-market fit? 

They asked every new customer one question after they've experienced the product's value. "How would you feel if you could no longer use the product?" Then, they measure the percentage of users who answer "very disappointed." If the answer is more than 40%, you have product-market fit.

It's important to note that neither Sean nor Rahul declared this as the only metric to measure a business's health. They also didn't restrict its use to a static point-in-time measurement. For example, Rahul said this in the same First Round Review article:

That's because the descriptions of product/market fit I found were immensely helpful for companies post-launch. If, after launch, revenue isn't growing, raising money is tough, the press doesn't want to talk to you and user growth is anemic, then you can safely conclude you don't have product/market fit. But in practice, because of my previous success as a founder, we didn't have problems raising money. We could have gotten press, but we were actively avoiding it. And user growth wasn't happening because we deliberately choosing not to onboard more users. We were pre-launch — and we didn't have any indicators to clearly illustrate our situation.

Sean Ellis said this in his description of the concept:

Admittedly this threshold is a bit arbitrary, but I defined it after comparing results across nearly 100 startups. Those that struggle for traction are always under 40%, while most that gain strong traction exceed 40%. Of course progressing beyond "early traction" requires that these users represent a large enough target market to build an interesting business.

It seems clear from these statements that neither meant to position the metric as the end all be all for business success. They also didn't indicate that this measurement should stop after reaching a specific destination.

The ultimate measurement of product-market fit comes down to revenue growth and high retention on a unit and revenue basis. I wrote a bit about these metrics here. The problem with these metrics is that they are all lagging indicators. When you have reliable net dollar retention numbers at scale, it's months or years from when you spent marketing dollars to attract and acquire the said customers. 

In contrast, the measurement that Sean and Rahul popularized is a leading indicator. It predicts retention.

Product-market fit isn't a destination. It's a journey.

The fascinating thing about products and startups is that they can gain or lose product-market fit at different points of time in their journey. 

I've seen this happen at every startup I've had the opportunity to explore closely - Padlet, Dropbox, Magento, Pilot, Invoice2go. Each company went through a journey of finding product-market fit. It was never a done deal. The understanding and the pull from the market evolved. The product evolved. 

Padlet is probably the most interesting one. It started as a friend's final year thesis project in ~2008 but sat on a server while the founder was working at other tech startups. He kept doing guerilla marketing to see if the product could find the ever-so-elusive concept of product-market fit. I got a call from him in 2012 that server traffic was blowing up—four years after the initial launch. The pull was from teachers/students. That market wasn't a conscious choice on the founder's part. Even after I jumped in as a co-founder and we started growing the product proactively, we didn't consciously define the market narrowly enough, even though, in hindsight, it was obvious where we had product-market fit. Also, just because we found product-market fit, it didn't mean we'd always have it. We struggled as we iterated on the feature set, evolved pricing, and go-to-market strategy. Product-market fit was a constant optimization. Fast forward to today, and it's a thriving business.

Dropbox is a beautiful example. As it expanded from its personal/consumer offering to a team/business offering, the initial product offering was a simple combination of the same individual/consumer offering with the obvious enterprise admin controls you'd expect to see in a business product. It wasn't for a year or two after the original launch of the team/enterprise offering that the picture became clear that we hadn't yet achieved product-market fit. While there was a strong pull from the market, the pitch wasn't resonating. The product wasn't ready. Reams of customer research, analysis of why businesses bought and churned, and sales feedback led the company to increase its investment in the business product to find proper product-market fit. Now, Dropbox's team offerings account for more than 35% of paid users.

Magento is another excellent example. It started and thrived as an open-source on-prem commerce platform. At its peak, there were more than 300k ecommerce sites powered by Magento. There was no doubt that it had product-market fit. Then with the advent of cloud and saas, things changed. The number of open source sites declined - replaced by other low-cost saas alternatives. With the market and customer needs evolving, the team had to reevaluate product/market fit. Over the years, the company had to relaunch/pivot the product to meet the market's demands - a cloud-based saas version. Now, that part of the business is a majority of new revenue. 

These are just three examples from my personal experience. There are so many such stories in the startup world. Every large/successful company has had to evolve in pursuit of the ever elusive concept of product-market fit. Either to support its core business or the introduction of new products. It's a constant evolution.

I believe this is what Rand's talking about - evolution. I agree with him. I think measuring a business's health is an ever-evolving process and journey. Not a destination. Rand's characterization of p/m fit as a destination is a misinterpretation. Or perhaps a gap in the explanation from our predecessors like Sean, Rahul, and Lenny. There's more agreement than disagreement from all sides. We're just getting stuck in semantics.

How do you measure it if it's constantly evolving?

It's pretty straightforward. To measure product-market fit, I recommend setting up an always-on measurement. I'm a big fan of Sprig, so I recommend you use that or something like it. I recommend asking the same question Sean and Rahul popularized. I'd encourage you to ask this question 30 to 45 days after customers experience a majority of the value from your product. This might be after their first payment or after you've completed onboarding.

Question: How would you feel if you could no longer use the product? 

Answers - Very disappointed, Somewhat disappointed, Not disappointed

Follow-up open text questions:

  • "How could xyz be improved to better meet your needs?"
  • "What are the primary benefits you receive from xyz?"
  • "Who do you think would benefit most from xyz?"

Now, measure the percentage of users who answer "very disappointed." If the answer is more than 40%, you have product-market fit. Remember, don't stop after that point. You need to measure this as you scale up constantly. The type of people you attract to your funnel might change. The product and first-time user experience might vary; hence, the metric above might move surprisingly.

Some more best practices as you implement this.

Target a big enough sample size. Qualtrics has a handy calculator to assess sample size. Say you're acquiring 5,000 new users a month. At a 95% confidence level and an error rate of +/-5%, you'll need 357 responses to have a statistically significant measure. If your population is smaller, say 100s, you'll notice that your sample size increases.

Increase response rates. Early on in your journey, you'll need to be quite aggressive to get a response from your users. If you're doing this in-product, I highly recommend designing this survey experience as a takeover of the entire UI. Reduce distraction to get higher click-through and completion rates. Don't rely just on email since there's a natural drop-off between email send, open, and click-through rates. If you're not getting enough responses, that might indicate an issue with product-market fit. If you have human touches with your customers, make this part of your onboarding or CS touch.

The real work begins when you have this data. Analyze it multiple ways to determine your product-market fit and where you don't. Are there segments where the response is higher? What's the total addressable market for that segment? Can that become the sole focus for you? Which segments are reacting poorly? Is there a pattern of needs you're not solving for them, and can you? Constantly refine the ideal customer profile definition based on what you're learning.

Customer Adoption Spectrum - an alternative?

Rand's alternative to product-market fit is the Customer Adoption Spectrum (CAS). CAS is very similar to the product-market fit methodology I've described here.

This visualization says that you have strong product-market fit with group B, and you still have more work to do in the other groups. Either you give up on those other groups because group B is big enough for now, or you need to evolve your product positioning/messaging to better attract and convert those other groups. 

While I love this view, there is one specific reason I believe this framework is less potent than the product-market fit methodology. Conversion metrics aren't the end all be all. They're not always perfectly correlated with retention. You might even find conversion rates to be inversely correlated with retention rates. This is because marketing and growth teams have become so obsessed with conversion and gotten so good at it that we're able to optimize conversion rates to a tee, all the while retention gets worse.

Rand's clear about how CAS is more adaptable, and you can swap out metrics. Great, so if you swap out conversion for retention, this becomes interesting. However, the challenge remains - retention is a lagging indicator. Ideally, you want a leading indicator.

And, we're back to product-market fit or some equivalent leading indicator that helps you illuminate the path towards scale. This is why I strongly believe in product-market fit being the better fit.

So, what's the verdict? Is Rand right or wrong?

Rand's right that if interpreted as a destination, product-market fit is not well suited to capturing the evolution of businesses. Having said that, I can't entirely agree with Rand's assertion that product-market fit is designed/meant to be interpreted as a destination. 

As you can see, in most business jargon, the discussion can get lost in absolutes. Truth is not absolute and is in the nuance. We all can do more to pick apart the concepts here. Rand could've taken the idea of product-market fit and built on top of it. Instead, he chose to create a different framework. There's nothing wrong with that; there's a ton of similarity between the two concepts.

The real lesson is that regardless of the framework, businesses can avoid resting on their laurels. Like Jeff Bezos says for Amazon, in business, every day is day 1. Keep figuring out whether your business is healthy and what you need to do to get or maintain your health. 

Now that you understand the difference between treating product-market fit as a destination vs. a journey, I hope you'll start measuring and tracking this critical metric regardless of your business's stage.

If you liked what you read here and are interested in building your go-to-market skills, follow GTM Digest on Twitter, where I post tips daily on leading go-to-market teams.