Too Many Damn Articles on Product-Market-Fit

Gil Rosen
14 min readDec 14, 2020

By Gil Rosen

What is product market fit?

How do you know when you have it?

Why does it even matter?

I do not think it means what you think it means

I thought to write an article on Product-Market-Fit based on my own experience as an entrepreneur, investor, advisor, and founder of A quick google found 17 articles on this very topic — half semi-vague, half contradictory, some relevant yet incomplete, and most highly redundant. So I decided to write yet another, synthesizing my own thoughts with relevant insight from those very articles (tldr and links at the end). But you’ll like this one, I promise. And when you’re done you’ll hopefully know what (I think) PMF is, how you know if you have it, and how it should affect your venture’s priorities.

Why should I, or you, even care? Because people make investment, partnership, sales, acquisition, and countless other decisions based on having it or not, and if you’re wrong, you might just make the wrong decision. And making wrong decisions can hurt. I say this from experience as I lived through it with one of my first investments whom I worked closely with.

They were a B2B software venture, let’s call them MIMFO (that’s not their real name, but MIMFO rhymes with nympho and they definitely screwed things up for a bit). MIMFO seemed to have everything going for them — multimillion dollar revenues, nearly 100 employees, marquee customers. They were convinced they had PMF and raised financing, kicked off campaigns, opened sales offices, and made strategic acquisitions based on this premise — yet when the rubber met the road they failed to sell their product beyond their first dozen customers. It was a great product, their customers loved it, it was timely and innovative, and yet they struggled. They missed targets. They had to lay off staff. Eventually, management was replaced. Why? They tried to scale before actually having PMF, which meant they burned cash on an assumption of customer value that wasn’t actually there. The first action the new CEO took was to determine how to get to PMF.

We’ll dive more into MIMFO later in this piece, but the lesson is to beware of scaling prematurely. A venture’s longevity is determined by their cash flow, and as soon as it begins to hire and scale, its burn rate increases. If a venture can’t show revenues or progress quickly, their cash runs out and they’ll find it hard to raise more. It is thus critical to scale only once PMF is actually found.

Product Market Fit?

So what is product market fit then? It is NOT as simple as “A product that fits a market”. Through my own experience as an entrepreneur, advisor, and investor in dozens of ventures, as well as my graduate studies and research, I’ve refined the definition I use to the following:

“A product, business model, and customer engagement method that consistently meets the needs of a specific customer segment or persona better than the alternatives, such that they’re willing to pay for it.”

I had the pleasure of learning about Product-Market-Fit from Andy Rachleff, CEO of Wealthfront, and co-founder of Benchmark Capital, in his course on Aligning Startups with their Markets at Stanford’s Graduate School of Business. Rachleff is credited with coining the term, but from his telling he was heavily influenced by Don Valentine of Sequoia as well as the lean startup ideologies of Steve Blank and Eric Reis.

  • Rachleff describes a company with PMF as one that has proven its Value Hypothesis; why a customer is likely to use a given product. Rachleff continues that knowing the value hypothesis identifies the requisite features, audience, and business model.
  • Steve Blank describes the criticality of both Value and Growth Hypotheses as testing whether a product or service delivers value to customers using it, and how new customers will discover a product or service, respectively.
  • Eric Reis describes PMF as when a widespread set of customers resonate with the product.
  • Alex Schultz, CMO of Facebook, describes PMF as nonzero retention over time
  • Marc Andreeson of a16z describes PMF as being in a good market with a product that can satisfy that market. This means being able to answer the “why”: problem/need, the “how”: business model, and the “what”: product/service

While these definitions make sense, their qualitative nature doesn’t help quantify what PMF actually is and how you know if you have it.

When asked for indicative benchmarks, Rachleff relayed

  • For consumer focused companies, PMF is generally found when you have organic viral growth. That is, people love your product enough that they use it, and then engage their friends to use it, who then engage their friends to use it.
  • For B2B focused companies, PMF is generally found when you have 7–10 repeatable sales to customers similar to each other.

Alex Schultz in a separate lecture at Stanford, details

  • If retention over time (how many customers remain from time of acquisition) asymptotes to a percentage common to a given industry, it’s indicative of PMF. Ie if over time retention stabilizes to a reasonable number for an industry; this might be ~30% for an eCommerce company, or ~80% for a social media company.

This last point is a critical addition, as high churn shortly after customer acquisition implies that the customer value isn’t actually there — be it for a dozen successful enterprise sales or exponential viral growth.

Andrew Chen of a16z describes metrics of PMF for SaaS companies as 5% free to paid conversion, 3x LTV to CAC ratio, and <2% churn.

While this is just an example, having a viable business model means having a product whose revenues will be meaningfully greater than the cost of sale. If a sale requires months of work for an expensive account executive but only yields $10K in revenue it’s likely not a viable business model — but if the annual subscription is >$100K or if no account executive is needed and users can learn of and download the product on their own, then it may be.

SaaS metrics of PMF such as NPS, and questions such as “would you be very disappointed if this product were to be discontinued” are great indicators of value, potential growth, and potential churn, but nothing beats actual numbers of true growth and churn from paying customers.

Using the above descriptions we can begin to triangulate on what PMF actually IS by also considering what it’s used for.

Why does PMF Matter?

The Rachleff/Blank/Reis camp believes a venture should prove its Value Hypothesis before testing and investing in its Growth Hypothesis, as it makes little sense to invest in attracting customers that either won’t convert into, or won’t remain, paying customers. Testing and proving growth hypotheses involves significant marketing and sales efforts, this means truly understanding a customer segment/persona, their needs, the value you’re creating for them and how to engage with them.

The idea of hypothesis testing is critical here and Rachleff provides insight that has served me well. Ventures, by definition, have limited resources. This means you can’t test every customer segment or every business model simultaneously. The value propositions and features required by different customer segments will necessarily be different and you only have so many engineering hours. The stories that resonate with different customer segments will be different. The channels they use will be different. Their buying personas will be different. And if you try to build and test for them simultaneously you can’t differentiate between a failed hypothesis and between a shoddy effort. Giving someone half of their requisite features likely won’t solve their true need. Sending out one email blast to six different segments will yield little return as the average prospect needs >8 touchpoints for conversion. 100 impressions on facebook for 10 different personas won’t give you as much valuable information as 1000 impressions for one persona, which is a large enough sample size to understand if an audience does or doesn’t resonate with a product and its messaging.

Where investors love to diversify, operators necessarily need focus. It is critical to come up with tests that will allow for minimal investment, yet still yield ample information on whether a value hypothesis has succeeded or failed.

By understanding the ramifications of PMF on the actions and investments a venture subsequently takes, we can understand the insufficiency of solely having exponential growth, low churn, or a dozen sales to enterprise customers. Rather, we need exponential growth for a well understood market segment — what pain or need is common across those customers such that we can build an effective marketing campaign that will resonate with them and reach them on fewer channels. We need enterprise sales to customers that are similar to one another, so that we can confirm our value proposition and ability to sell is repeatable and predictable. We need to understand which customer personas have low or high churn to invest in features for those that will remain customers.

Taking all of this into consideration yields the definition provided earlier:

“A product, business model, and customer engagement method that consistently meets the needs of a specific customer segment or persona better than the alternatives, such that they’re willing to pay for it.”

PMF Broken Down

Let’s unpack this a bit.

PMF denotes:

  • There is a concrete identifiable customer segment or persona in focus
  • We understand their specific needs and pains
  • We have a product AND business model which address those pains
  • We understand how to communicate and engage with these customers so they understand and appreciate our value proposition
  • They care enough about our solution meeting their needs that they’re willing to pay for it — either directly, or indirectly (eg advertising, information, etc)

The above does not imply that you have a successful company, as perhaps the market you have found is small or dying. Nor does it imply that one should start with a customer segment/market and then define a product — that is an entire debate in and of itself which we’ll save for another article. Rather this merely indicates that a venture has found PMF for a specific product in a specific and clearly identifiable market.

The metrics around NPS and organic viral growth indicate that a product or service has evident enough value that people can recognize it, use it, pay for it, continue to use it, and recommend it to the peers they feel would benefit from it. For this to happen the product must really speak to their common yet specific needs.

The metrics around multiple sales to a similar customer segment indicate that multiple customers in that segment share in the same challenge, that your team is able to communicate that challenge and your value proposition, and that multiple customers in that segment find your product or service as solving that challenge to a degree they’re willing to pay for. One or two sales aren’t enough because perhaps you’re solving a fringe challenge that isn’t actually common for the segment, but if you get to 7–10 customers, it’s highly indicative that you’re on to something.

Once we are in a position where we are indeed clear and confident in the ability for our product/service and business model to meet the needs of a specific market, and our ability to understand and communicate with that market, we can begin to test our growth hypotheses and invest in scale.

What could go wrong?

There are countless stories of companies that lacked one or more of the PMF criteria, scaled early, and then struggled — some pivoting successfully, others less so. I’ve analyzed a few here to read as they interest you.

Dropbox found viral growth quickly, yet when it tried to monetize its consumer offering people switched to free alternatives such as Google drive. Most consumers weren’t willing to pay for the value Dropbox was offering. Ultimately Dropbox pivoted to focus on enterprise vs consumer needs and was able to profitably scale, but this is a clear example of over-indexing on one metric (growth and users) without considering the bigger picture — value people are willing to pay for.

This doesn’t mean all companies need show early revenues either — Facebook and Snapchat are two examples of companies that waited to focus on revenues, but they understood that their value was held in user data that could be monetized for hyper-focused advertising. They continued to focus on user value and growth, all the while aggregating valuable user data they believed would translate into advertising revenue.

C3IoT, the IoT platform started by Tom Siebel tried to circumvent the usual playbook by using Siebel’s deep financial resources and connections to sell their product to customers across multiple industries. Yet ultimately the engineering and customer support resources required to meet the needs of disparate customers across disparate industries with different priorities proved intractable at such an early stage, and they scaled back to focus on the Energy space.

Indeed this was the issue MIMFO faced as well. The CEO and COO were highly intelligent and highly technical problem solvers. They were able to successfully sell their software to a dozen companies in their geography across the Media, Energy, Finance, and Telecom verticals. Their customers were happy. They and their investors were sure they were ready for scale. Yet a closer look indicated that it was only the founders that were successfully selling because they could confidently tailor the product to customers’ needs. They would then engage product and professional services to build what they had promised. But this approach is not sustainable. You can’t expect your average sales person to reinvent the product for every customer, and building bespoke products presents tremendous challenges when you scale and need to meet the disparate product needs of hundreds of customers across a dozen customer segments. Once the company had exhausted their immediate network of prospects, they struggled to convert marketing leads into sales leads because they didn’t have a clear value proposition for a given customer persona. What they had actually built was a software enabled services company, which is indeed a viable business, but a very different one than a B2B software business. Palantir is one such example of a successful software enabled professional services business, and while it has found success, most Venture Capitalists are not interested in professional services heavy ventures, as scaling people is notoriously difficult and professional services margins (and valuation multiples) are significantly lower than software. After multiple quarters of missed targets, the board replaced the management team and refocused the company on the consumer banking vertical, righting the ship.

Groupon is famed for having grown virally and even IPO’d without truly having found a compelling value proposition for its paying clientele — the vendors selling products/services. Groupon promised to attract new customers to vendors with compelling deals, often sold at a loss for the vendor. While people indeed took advantage of the deals, there was no guarantee they weren’t already existing customers, nor that they would remain customers at the regular price. In fact, with competing services on Groupon, a customer could simply go from deal to deal and never remain loyal or pay full price. Moreover while there are industries where Groupon provided value, eg selling outstanding inventory of perishable/expiring goods, be they food, flights, or hotel rooms, for many industries it simply exposed businesses to customer segments that weren’t in their price category to begin with and quickly churned.

Succeeding in customer growth without a viable business model has been a recurring theme in meal-kit delivery services such as Sprig, Chef’d, and Munchery. They addressed a real problem, cooking regular meals, but could not do so profitably given the high costs of, well, everything, especially delivery. That coupled with a low barrier to entry makes it difficult to succeed from the getgo — perhaps an Uber or Postmates that have a reusable delivery infrastructure, or grocery retailers that have cold storage chains and access to raw materials and distribution, would fare better. Indeed, Blue Apron tackled high delivery costs by using the mail and is finally showing profitability post-pandemic.

One company I invested in focused on CPG planning software. They had a strong product that saved millions in logistics and production costs, and multiple big brands as successful customers. Yet the value was distributed across multiple disparate orgs, and the sales and implementation cycles were complex and lengthy, making it challenging to find a customer champion to push the sale through and ensure they could realize its value. The product was great, the market was large, the value was there, but engaging the customer in a manner to realize that value proved challenging.

Another two of my portfolio companies Oxygen and Bellhop found success after multiple pivots finding their market fit both in segment and in business model. started as a banking platform for the gig economy — banking and credit facility are notoriously difficult for those not on W2 with proof of regular salary. Oxygen initially focused on the Uber/Postmates/Taskrabbit driver segment and saw customer growth offering banking and credit services. Yet their possible value propositions were limited when compared to the freelancer/creative segment where they could offer both personal and business banking services including invoicing, budget management, and cash flow projections for the digital-first generation, while monetizing card transactions and credit.

Bellhop began as an aggregator for travel services — a one stop shop for the local uber/doordash/opentable. It quickly streamlined its focus to just ridesharing — providing users with options to choose between Uber/Lyft/Curb/Via etc. optimizing for price or speed. While it saw growth, margins remained tight as Lyft/Uber had the lion’s share of rides. Lyft and Uber’s loyalty programs and drivers operating on multiple platforms began to eat away at both retention and value. Bellhop has since expanded in two critical ways 1) onboarding legacy black car companies and taxi fleets as a monetizable distribution channel for their services; true airport pickup for city taxis, limo/van/scheduled pickup/long trip for the black car companies. 2)aggregation for bikes and scooters which cannot come to you and where closest proximity is advantageous. Moreover, Bellhop uniquely has aggregate user mobility data of value to utilities in planning for an electric future, or urban transport planners. These changes both enable a singular focus on the mobility challenge for a user — how to get from point A to point B, and how to better monetize the traditionally low margins of ridesharing to build a more profitable business model.

Go forth and fit

So, to summarize, Product-Market-Fit describes:

A product, business model, and customer engagement method that consistently meets the needs of a specific customer segment or persona better than the alternatives, such that they’re willing to pay for it.

This means:

  • There is a concrete identifiable customer segment or persona in focus
  • We understand their specific needs and pains
  • We have a product AND business model which address those pains
  • We understand how to communicate and engage with these customers so they understand and appreciate our value proposition
  • They must care enough about our solution meeting their needs that they’re willing to pay for it — either directly, or indirectly (eg advertising, information, etc).

Indicative benchmarks for PMF are:

  • For consumer focused companies — organic viral growth. That is, people love your product enough that they use it, and then engage their friends to use it, who then engage their friends to use it.
  • For B2B focused companies, when you have 7–10 repeatable sales to customers similar to each other.
  • Reasonable customer retention for the relevant industry (can find industry benchmarks)

Until a venture team finds PMF they should likely remain lean and test hypotheses for different product/market/business model combinations sequentially, investing enough in each to know as quickly as possible if they’re viable or not, and moving on to the next one if not. Once there’s enough confidence in PMF, they should invest in hiring the right personnel to test and execute the growth hypothesis for how to best scale the company.

Assuming PMF before you’re there is the surest way to burn cash and crash.


“Aligning Startups With Their Market” — Stanford GSB course notes and content



Gil Rosen

Gil is a serial entrepreneur and early stage investor, advisor, and founder of