3 Principles for Assessing Product-Market Fit

I was recently listening to 20VC founder Harry Stebbings’ podcast interview with DoorDash VP Brian Hale and it got me thinking about the importance of customer retention as it relates to projecting business success.

While I agree with most of what Stebbings and Hale had to say (the episode is worth a listen), there are some additional nuances that our team at Headline uses to evaluate product-market fit for companies. Without this insight, investors can’t get the full picture of how a prospect may perform long-term. And in increasingly volatile market conditions, a clear definition of product-market fit is more important than ever for founders and VCs. This year’s sustained slowdown in venture capital funding likely has many startup founders on edge. Amid inflation, rate hikes, increasingly spending-conscious consumers and general market volatility, it’s a challenging time to grow a business. 

But that doesn’t mean it’s impossible—far from it. Finding success means doubling-down on business fundamentals. I’ve written previously about capital efficiency, and product-market fit is another, equally important, metric for founders and VCs to drive toward in order to achieve growth no matter the market conditions.

Pinpointing product-market fit

When a company has product-market fit, its product or service is positioned to a market segment that values it, which means there’s a sustainable demand for it. Even the most ingenious ideas can’t get a foothold when customers aren’t willing to make them part of their lives over the long-term. Product-market fit is what all VCs are looking for because it’s critical for building a large, successful company. Once there’s product-market fit, it’s simply a matter of getting the product or service in front of the target customer base.

At Headline, we focus heavily on product-market fit. I spend huge chunks of my day reviewing company data looking for signals about product-market fit to understand whether our team should be leading funding in a given company, and what is special about a given company’s execution to date: Are we seeing outlier customer retention and engagement? How is this engagement changing (improving/deteriorating) as the business scaled over time?

Using our proprietary software, our team has developed a model that looks at three key components of product-market fit to make the best assessment of a company’s performance to date. By better understanding how to measure product-market fit, founders and VCs alike can better understand real growth opportunities—regardless of market conditions.

1. Assessing product-market fit starts with cohorts and extends to month over month retention

One aspect of Stebbings’ and Hale’s discussion on product-market fit centers around cohort analysis, which resonates with our view at Headline.

“Take everyone who started using [your product] in the month of March,” Hale says. “Now, let’s follow them forward in time…How many of them are still using this product on a regular basis in the following months? Often, it takes a long time to know whether you have product-market fit because you actually need enough time to see whether retention stabilizes. We call these J-curves: when you graph from the people who started, forward in time, and you see what percentage of that group is still around.” 

The traditional way to evaluate retention was to evaluate trends on a month-to-calendar month basis. However, this does not provide a clear picture of customer behavior (customers who have been with a product for over a year are very different from those who have just joined). As a more effective alternative, Hale highlights the importance of evaluating cohorts to understand customer behavior to see retention over time. 

Here’s a graph showing user retention by cohort:

• Each line represents a cohort’s retention over time. As churn within a cohort reduces, the lines flatten out.
  
• Time is measured by lifemonths (how many months a given customer cohort has been with a business) instead of calendar months to provide a more meaningful comparison across cohorts 

The flattening lines indicate product-market fit because they show a subset of customers in the cohorts continue to use the product or service long-term. This engagement indicates that the product or service has become part of those customers’ everyday lives. 

We also like to compare lifemonth-to-lifemonth retention, which compares the amount of customers that retain from one lifemonth to the next (instead of the opening amount of customers in a cohort). When the lifemonth to lifemonth reaches and sustains at 100%, that’s product-market fit:

Most VCs and operators are looking to see these lines flatten out. But the horizontal asymptote is just the beginning for our team at Headline. 

2. Cohort behavior over the long term indicates stronger product-market fit 

To further evaluate product-market fit, we also like to look at the behavior of the customers who retain long-term. When retained customers gain more value as they mature with a product or service—which can be seen by customers being more active or spending more with a business over time—that’s a stronger indicator of product-market fit. For these cohorts, the asymptotes of the cohort lines may become horizontal and then increase as the retained customers increase usage or purchases. 

Here’s an example:

When retention asymptotes flatten out and the usage of the retained customer increases, is a good indication that the product or service is hitting a real nerve in the marketplace. Every one of our successful portfolio companies fits this model, if it doesn’t we usually don’t invest.

3. Lower asymptotes don’t always mean product-market fit is zero

So what about companies that have very low (1-2%) retention among cohorts? Stebbings and Hale touch on this in their podcast, too. Hale indicates that the asymptote quantifies the level of product-market fit, and that retention with a low asymptote is the equivalent of having low or no product-market fit.  

Of course, retention with higher asymptotes is favorable. However, a low asymptote may not always be an indication of poor product-market fit. Instead, it could be a reflection of a wide customer acquisition funnel. For example, when there’s very little friction to start using a product or service such as with a freemium offering, we tend to see lower asymptotes (compared to expensive services such as enterprise software products that have $100,000 implementation fees).

If a freemium service has a low asymptote, this indicates that a small percentage of customers remain. It does not mean there’s poor or no product-market fit as it could be that the marketer is just incredibly good at getting customers to try the given product.

Because wide acquisition funnels with low friction can have lower asymptotes, we do not automatically disregard these start-ups since this doesn’t provide the full view of what’s happening as cohorts mature. It’s possible that companies with low retention asymptotes have solid product-market fit, just only with a subset of the customers they acquire. If so, we’d ask more questions such as: How expensive is it to acquire customers versus the customer lifetime value earned? Is there enough market size to sustain this low retention asymptote?

A business with a low retention needs to churn through a large number of customers to achieve a large scale. For example, a company with a 2% Retention asymptote needs to churn through 50 customers for each one that it keeps: 2% Retention = 1 out of 50 customers remain (49 churn out). Therefore, the question becomes whether the business would churn through its entire target market before it gets to a venture scale business ($1B+ public company).  

The answer to these questions can’t be found in these graphs. Further examination is needed but it’s important to note that start-ups with low retention shouldn’t automatically be dismissed as “leaky buckets” with no product-market fit.