A startup’s market size must be big enough to justify any upfront investment. Applying this logic suggests $5 million is a very minimum empirical market size for software startups. The same logic suggests very minimum market sizes for hardware should be larger, and, for pharma, much larger. Rigorously validating market size requires time and money and this imposes an upper limit on market size. Perhaps the only justifiable support for market size is sales (or pre-sales) because of the uncertainty of sales estimations for very early stage startup products. If sales (or pre-sales) are the only legitimate way of usefully calculating market size, then market sizes quoted by startups should be very small indeed, perhaps $100k to $1 million in size – much smaller than the $100 million or billion dollar market sizes often touted today.
To calculate market size, start with a figure of $1 billion per year. Then, project a price (let’s say $1, for this fictitious app). Then, divide the market size by the projected price to project the number of customers (1 billion, in this case). Then, if you feel the number of customers is too high, reduce the projected number of customers (say, to 100 million) and compensate by increasing the projected price (say, to $10). Iterate until you are happy. You now have a projected market size that you can back up with a projected price and a projected number of customers.
For the avoidance of any confusion, the above paragraph is a joke. However, as with many jokes, there is perhaps an element of truth behind it. Market size is a much talked about concept in entrepreneurship but also one that I feel is incredibly murky.
Anecdotally, there are many market size heuristics; for example, some investors want to see a billion dollar market and some some want $300 million in year 5 revenue. Some investors want the opposite; Peter Thiel, in Zero to One, for example, is scared by founders who project large markets and prefers to see companies start by building small monopolies. Bill Aulet, in Disciplined Entrepreneurship, also favours a small beach-head market and suggests initially aiming for markets that are between $5 million to $100 million in size (based on the startup’s maximum potential revenue). So, the question is, what is the ideal sized market? and why?
WHY DEFINE A MARKET SIZE?
On a simple level, you could say that you define a market size to see if you could make a boat-load of money – the more the better. That could be true, but I think there’s a more refined reason. Ultimately, if you’re a founder, you’re probably passionate about something and you want to bring that something to life. For it to come to life, one requirement is that the business is a sustainable one – market size, in my opinion is part of that consideration.
WHAT MAKES A MARKET TOO SMALL?
To get to a sustainable business there are some barriers to entry that need to be surpassed. For software, it might be a few years of salary for founders and employees, e.g. a million dollars. For hardware, there might be additional R&D costs, e.g. a few million more dollars. For pharma, there might be also be the cost of clinical trials, e.g. hundreds of millions of dollars. If your business is going to be successful, you probably want your eventual annual revenue (one way to define market size) to be significantly larger than this barrier to entry. This sets a minimum market size.
One way to think about this is to consider the final worth of your company as being the maximum achievable revenue (which we’ll consider to be the market size) multiplied by the net profit margin multiplied by a price to earnings ratio:
FINAL COMPANY WORTH = MARKET SIZE x PROFIT MARGIN x PRICE TO EARNINGS RATIO
Consider that the final company worth will need to be large enough to justify the capital required multiplied by the return on investment that investors require:
FINAL COMPANY WORTH = CAPITAL REQUIRED x MULTIPLE OF RETURN ON CAPITAL
This means that, the minimum market size you require for a sustainable business is:
MINIMUM MARKET SIZE = (CAPITAL REQUIRED x MULTIPLE OF RETURN ON CAPITAL) / (PROFIT MARGIN x PRICE TO EARNINGS RATIO)
To put this in context, we can consider a price-to-earnings-ratio of 6X (a possible value for mature private companies), a return-on-capital for investors of 20X and a profit margin of 50%. This would give the following minimum annual earnings:
- Software example: Minimum Capital Required = $1 million; Minimum Market Size = $6.5 million
- Hardware example: Minimum Capital Required = $3 million; Minimum Market Size = $20 million
- Pharma example: Minimum Capital Required = $200 million; Minimum Market Size = $1.3 billion
Obviously, the minimum capital required depends on the startup in question, as well as other assumed values above. However, I think the above calculation provides some good heuristics:
- Markets smaller than roughly $5 million (measured in achievable revenue) probably don’t make much sense to pursue.
- The more capital you think you need to bring your business to life, the larger the market size you will require.
WHAT MAKES A MARKET TOO BIG?
In my opinion, the goal of calculating market size is to demonstrate that sufficient revenue can be generated to justify the upfront investment that is required to bring the startup to life. There are two ways you could do this: extrapolation or validation.
Extrapolation – here, you take whatever insights you have from a first sale (hopefully you have one) and try to project future revenues (preferably using outside data sets of relevance). It isn’t often stated explicitly but this kind of projection is very difficult because there are many reasons why current and future customers might be different. For example, the problem you solve now may not be a priority for new customers, new customers might already have an alternative solution, or, it may be more difficult to reach new customers. As an early stage startup I think the ability to extrapolate to new customers is very limited. Before a product is fully defined, the business model repeatedly tested and the sales cycle fully validated, I would ask whether projecting revenues is a waste of time.
Validation – here, you sell (or pre-sell) your product to a customer (or group of customers) and get an idea of your revenue and perhaps also the rough cost of acquiring the customer (i.e. marketing + sales costs). Getting validation like this isn’t easy but it does allow you to establish a minimum market size from the revenue. From this starting point, you can maybe also include in the market size any customers are extremely similar (if not identical). However, from here, you hit a trade-off. By doing more validation you can increase your market size. However, validation work costs time and/or money and a point is reached where you stop, look at the current market size, and see if justifies an investment that will accelerate market validation.
Ultimately, market size validation is, by nature, an inefficient process and its expense imposes a practical limit on market size. First of all, founders can’t outsource or easily automate the validation process because it is crucial to build relationships with customers and better understand their needs for themselves. This means that the cost of acquiring customers is the founder’s opportunity cost (i.e. the salary foregone in choosing not to work for another company). This could be well above $100k per founder per year. The second reason why validation is expensive is because it is essentially a low efficiency direct sales process. Mature startups have a more clearly defined product, a tested sales funnel and low customer acquisition costs. Early stage startups are still figuring things out and this makes sales/validation less efficient and more expensive. For example, while mature startups might pay $1 to acquire every $5 of future customer value, during validation, startups might effectively spend $1 on direct sales to acquire only $1, or even less, in future customer value. With this, admittedly blackbox, assumption for validation, a founder whose opportunity cost is $100k per year would validate a market size of $100k, which doesn’t even meet our minimum market size requirement of $5 million.
My model of market validation is almost surely wrong. However, I think it does illustrate how expensive it is for founders to properly validate market size. Clearly the idea that it is possible to validate, or even estimate with a useful degree of confidence, a market that is $1 billion in size is absolutely ludicrous. Validating even a few million dollars of revenue, I think, is something that requires extreme founder sales talent and skill.
I think that market size is a concept that needs serious reconsideration. I worry that founders and entrepreneurs are wasting significant time in performing and analysing market size estimations. I wonder whether, and hypothesise that, current methods are more likely to lead everyone astray than in the right direction.
Clearly it is important to know the market size because the business can’t be sustainable if the market isn’t bigger than the investment that is required upfront. I think it is important to question whether, for startups, it is possible to make projections of market size that include customers to whom the product has not yet been sold. Sales (or pre-sales), I think, are a much more sure way to assess market size. Even if the market size is then only hundreds of thousands or millions, I wonder whether having that small quantity of certain information is better than having a market size of billions based on useless information.