The Smallest Market I’ll Go After is $200m

I recently decided to not invest any more in Futurecam, pivoting from a product company to a consultancy.

One of the reasons for this decision is that the market for third-party iPhone camera apps is too small.

Let me walk you through how I arrived at this conclusion, because this is a useful filter to apply before starting a startup:

A rule of thumb in business is that one can capture only 1–5% of a market. 1%, if the product doesn’t have network effects [1].

Let’s look at this 1% number in more detail. I won’t be able to reach out to all my target users. Say I’m able to reach out to a third of them. Of them, only some will actually turn to be in the target market, say two-thirds. Of them, only some will bother to install the app or sign up for the service, say a third. Of them, only some will pay, say a third. Multiplying all these percentages together — 33% x 67% x 33% x 33% — results in 2%. Not far off from the 1%. Let’s stick with 1%, since it’s more conservative, and it’s a nice round number for quick mental calculations.

So, let’s say I’ll have a 1% market share. Now the question is: does that generate enough money for me?

The company should have an annual profit of at least $1m, to pay me and a marketing / sales cofounder or employee $500K each. I don’t have good enough marketing / sales skills, so I need to account for another person. But why do I want $500K? Because I can easily get a job as an employee that pays me $100K a year. Since being a founder is much riskier, I want $500K to compensate me for that risk. Otherwise, it will be a bad decision for me to start a startup.

To make an annual profit of $1m, we need a revenue of say $2m, accounting for taxes, Apple or other gatekeepers stealing 30% of your revenue, a coworking space, laptops, and lots of other overheads that may be small individually, but are significant in aggregate.

Since we’re assuming 1% market share, this means I need to target a market that’s at least $200m.

When I say $200m, the order of magnitude is the key part, not the precise number like whether it’s $100m or $300m. $20m is too small. And a billion dollars is probably too high a threshold — it eliminates a lot of niche businesses that may be blue oceans because others have overlooked them.

So, $200m it is.

You can change this calculation as you want. Take your market-rate salary, and multiply it by 5 to compensate you for the huge risk in running a startup. Then double that for your partner. Double that again to account for costs. Then multiply that by 100, assuming you’ll capture 1% of the market. Feel free to change all these factors as you think is appropriate. The goal of this blog post is not just to give you the conclusion but also, if you disagree with the $200m number, to give you a framework to calculate your own conclusion based on your circumstances and assumptions.

Keep your assumptions conservative. We’re often tempted and put in bigger numbers because we’re excited about an idea and are unknowingly biased about its potential. Don’t do that. And it’s better to be pleasantly surprised later on by making a bigger profit than expected than to be unpleasantly surprised.

Use the market size threshold as a filter to weed out markets that are too small, and don’t even enter them. You don’t want to toil for years and fail because the fundamentals were wrong in the first place.

Businesses Can’t Scale Down Below a Threshold

In theory, you can target as small a market as you want, as long as your costs in running that business are proportionally lower. Including cost in your time.

But, in practice, this doesn’t work, for a variety of reasons:

First, you need a high level of competence to make any product successful. Then you need to make the business successful. Being a great entry-level engineer isn’t enough. Being a great engineer isn’t enough, either. You need skills in different areas like UX design, project management, marketing, sales and business.

Second, if you have multiple such skills, you’re typically a senior, highly skilled professional, so you have a high opportunity cost. For example, I can easily get a job as a CTO of a startup, so for me to start a startup, it should have a high enough value.

Third, these skills are typically not all found in one person, so you’re talking about a partner, whether cofounder or employee, which means the market should be twice as big, to pay enough for two people.

Fourth, if a market is small enough to make it viable for me to work only part-time, I wouldn’t consider it. It’s fine to start part-time, but not if that’s all the market can support. There’s a context-switching overhead, so you won’t be productive with two half-time jobs, certainly not as you would a full-time job. Not to mention that having a side gig can come in the way of getting a full-time job. Or working full-time on some other idea you come up with in the future.

Fifth, if you have too small a market, in theory you can compensate for it by getting a bigger piece of the pie. After all, smaller markets typically have fewer players, so each should be able to get a bigger piece of the pie. In the limit case, a tiny market is enough if you can get all of it. But that’s all in theory. In practice, you can’t get 100% of a market. Identifying and connecting to potential users is hard. Ignore the hype about the Internet removing barriers. If anything, the barriers are becoming higher as there are more and more products competing for limited attention of users. Paul Graham says that most startups fail not because of their competitors. I found this to be true with Futurecam, too: I don’t think many potential users looked at it as a choice between Futurecam and one of its competitors. Instead, the choice was whether to use Futurecam. So, reaching out to and convincing potential users to become actual, paying users is hard irrespective of whether there are competitors. If you can reach out and convert all the users in a given market, if your marketing skills are that good, everyone will line up outside your door to get you on board as a marketing cofounder or CMO, and you wouldn’t be considering such a niche market! In practice, you’ll reach out to only some potential users. You’ll then find that only some of them are actually in your target audience. Only some of them will bother to install your app. Only some of them will pay. If you start with a small market, and further reduce it multiple times, you’ll have only enough money for a coffee.

1000 True Fans? Not Enough

A fascinating exploration of how small a company can get comes from the idea that it’s enough to have 1000 true fans. You should read this article, and I like its first principles thinking, but I don’t agree with it for two reasons:

• It assumes I want to make only $100K, but I want to make $500K as a founder, to compensate me for the huge risk involved.
• It assumes each user, or “fan” as the article calls it, is willing to pay me $100 a year. That’s too optimistic. First, that’s a lot of money, especially outside the US. Second, to get paid that much, I’ll need to build a sophisticated product, which means a bigger team, which means that the assumption of $1m profit we started out with is no longer sufficient. Third, I don’t know how to get $100 ARPU; I know how to get a dollar or two ARPU. Fourth, if your product or service is that good, there’s a chance you’re exploiting a temporary anomaly in the market and it won’t last. The only service I paid $100 a year to was Dropbox, and that too only for 2–3 years before Google Drive launched. Now I pay OneDrive $23 for the same 100GB of storage. $100 is unsustainable. Fifth, the word “fan” implies an irrational liking for your thing, and most people are not fans. Most products have users, not fans, so we need to design for users.

So, going by what’s reasonable and what I know I can do, if the company should make $2m, and each user pays $1, we need 2 million users.

So the smallest market I’ll go after is $200m, and has 2 million users.

[1] If anything, this space is over-explored, since every VC and founder wants network effects.