On Growth in Early Stage Startups
Getting to Product-Market Fit
This article is 2nd in the ‘On Growth’ series. In the first article, we covered the objective and outline of the series. In this article, we will cover:
- What should growth function of an early stage startup be doing?
- Who owns the growth function of an early stage startup?
- How should an early stage startup grow? Are there common ways?
Let’s begin.
Q: What should we be doing? A: Getting to product-market fit, of course.
Getting to product-market fit (referred to as PMF) is the most, and kind of the only, important thing as you start up, goes the common advice. And it’s true.
Any business, at its very core, serves two purposes: value for buyer/user, and profit for seller/owner. The term PMF emphasizes the fundamental nature of this economic principle: both the components need to exist and they need to fit.
For example, if I start a confectionery store, I need to have sweet things to sell and there need to be enough people around who are ready to buy them in the quantity and at the price that my business remains feasible. That’s PMF.
But I have observed that most people get what PMF is and why it is foundational. What trips people up is how to measure it and to arrive at a binary yes/no answer for the question: does my business have PMF?
Qualitatively it has been described as: when the market is so strong that product is flying off the shelves; when you have to keep adding servers; and so on. But how do we measure it quantitatively?
Economically, if your business can make profit, it has PMF. Inversely, it can’t, it doesn’t have.
Let me explain in more detail.
In a setting like a confectionery store, I’d argue that profit on every transaction (with fixed costs like rent, cost of goods, etc., factored in every transaction) is the indicator of PMF.
In a technology-driven marketplace, the criterion of profitability still remains but instead of measuring it for every transaction, you can now measure it for a customer over the customer’s lifetime (typically 12 to 18 months from first transaction, depending on the nature of business). For example, if your online confectionery shop has 1000 customers who will, on an average, deliver you 1 dollar profit per customer, over their lifetime, your business is worth 1000 dollars (ignoring net present value calculation, for simplicity).
But that’s too long term and the formula for calculating customer lifetime value will have too many variables that would be unknown at the early stage of a startup that we are discussing here. So, you need a leading indicator.
That leading indicator is retention i.e. number of customers from a particular cohort (also called vintage in some industries — basically a set of people who transacted for the first time with your business during the same time period) returning to you. For example, if 100 customers came to my confectionery shop in first week of business, and out of those 100 customers, 10 came back next week too, the shop has a 10% retention in week-1 (or week-2, depending on the nomenclature you adopt, but it is common to label the first day/week/month/year of cohort as day-0/week-0/month-0/year-0).
Retention graph (with the % of active customers from the same cohort on y-axis and time in days/weeks on x-axis) has a sharp drop initially. The extent of the drop determines whether you have PMF or not. And since the drop comes very early (e.g. most consumer apps lose more than 80% of users on day-1), it makes for a perfect early indicator.
This calculation is extremely important to get right. My confectionery store might get 100 customers in week-1 and 80 customers in week-2, but that’s insufficient information for calculating retention. If all 80 customers in week-2 are new and none of the customers from week-1 returned, I have 0% retention and probably no PMF (or perhaps week is not the right time interval to buy sweet stuff again — I disagree but, hey, to each their own — and we should be measuring retention across months). However, if all 80 customers are from the past week, I have a great PMF and now can progress to the next stage of problem-solving i.e. getting new customers.
So, what’s a good retention % to have? Whatever makes your customer lifetime value greater than customer lifetime cost is a good retention %. But, for simplicity at the outset, you can look up the benchmark for your domain and geography, and try to beat it.
The secondary leading indicators for PMF are frequency and engagement: of the people coming back, how often are they coming back and how much are they using the product whenever they do come. Retention %, frequency of usage by retained users, and average revenue per usage — these are three mutually exclusive factors using which you should be able to get a good idea of whether your business has PMF or not, but often the biggest determinant is retention %.
If you find that you do not have PMF or a weak PMF, you should dig deep into how you can solve retention %. This is highly contextual to your business but you should start by talking to all the users of the cohort.
By the way, there are more caveats to this (is the cohort homogeneous, for example) and we’ll come back to this topic again in the next article in this series. But if an early stage startup is regularly getting more cohorts in and is measuring their retention, that’s good enough for this stage, I believe, and it is already way more than most startups do.
Q: Who should be doing it? A: CEO is the first (and last) Head of Growth
Ensuring profitability and growth of the business are the core jobs of its CEO. These jobs can eventually be delegated but will always be managed more closely than other jobs by the CEO.
In the initial stage, the founder-CEO typically is the only person managing growth. Once there are enough users to have significant enough data to do PMF calculations, and there is a small but steady inflow of users in every cohort, it’s not uncommon for a CEO to get the first member in the growth team to execute the tasks and complete the measurements.
Q: How should we do it? A: Growth Hacks, yes please
To look for initial demand (key job is still to measure PMF but you need to get significant enough data), one generally goes back to users whose pain points the product was supposed to be solving to begin with. For example, if you designed a tool for social media influencers, you would just go to them. If you were scratching your own itch, you will look at your peers as potential customers.
Remember when I said in the previous article that there are no growth hacks. I was lying. There are growth hacks. The name does trivialize the practice but you can be innovative at this stage to jump-start growth.
Over time, the probability of one innovation in distribution of your product making significant dent to your growth rate will decrease. But at this phase, whatever gets you to the next milestone — 100 customers, 1000 customers, and so on — is fine. PMF data becomes stronger and clearer with more cohorts being added.
There are a lot of anecdotes about how even to-be giants, like Airbnb and Uber, did innovations to solve the initial need for liquidity on the platform but thinking we will growth hack it or we will go viral like them are terrible ways to think about distribution (thinking in terms of anecdotes, in general, is bad). An innovation has a terribly short half-life and you have to start with your product and your market and then think about what you could be doing to distribute your product to the relevant market in a quick way.
Thankfully, this aspect of growth has been well written about by its practitioners. While it is hard to have an exhaustive list with all the innovation happening around, books like Traction by Gabriel Weinberg and Hacking Growth by Sean Ellis lay down the framework of most common ways. Lenny Rachitsky’s blog, among others, covers this topic in amazing depth too.
We have now covered key questions for Growth function in early stage companies that we set out to answer. Got a question or a comment? Please do share.
In the next article in this series, we will cover Growth function in post-PMF companies. This corresponds to the phase around and after Series A in startups that have raised funding by venture capital firms.
PS: This article is perhaps the weakest in the series, considering it’s the territory of founder-CEO, but I had to go in chronological order. I am excited about the next article and, hopefully, it has more original thoughts on that phase. See you there.
