I was listening to something this morning about data mining and it got me to thinking about data and entrepreneurship. Large companies and high tech startup have access to a wealth of data. They have so much that they pay big money to reduce that data down into more manageable, and thus actionable, chunks. They take thousands of impressions on a web page and determine what shade of blue the call to action button should be to get the most engagement. (Probably not blue at all, but you get the point.) What sizes should we make and how many? If you have a wealth of previous information or access to a firehose of potential customers to experiment on, those are trivial questions. I could go on, as could you.
But entrepreneurs, at least my entrepreneurs, don’t have data. They especially don’t have tens of thousands of data points on everything they need to know to make decisions, and even if they did they would lack the resources to make sense of it all. So what do they do?
Well, many entrepreneurs, and student entrepreneurs are no worse than entrepreneurs in the ‘real world’ about this, start with a sample size of one. They assume that if they want something, everyone wants it. As a consultant to entrepreneurs, student and otherwise, I talk them down from this ledge. I used to encourage them to survey at least 50 people in their target market to determine anything important. I knew in my heart of hearts that that wasn’t enough, but it’s what I was required to do in one of the premier entrepreneurship programs in the country, and so I reasoned, it was at least a good start. It’s hard. It’s uncomfortable. You quite often don’t get the answers you want. But if you find out you’re wrong before you spend a lot of money, a loss is a win. This is the heart of the lean startup tools used outside the digital space, such as the business model canvas.
The business model canvas approach relies on validating assumptions, and assumes a relatively small sample size is adequate to validate an assumption. How small depends on the question. A question of logistics can sometimes be answered with a sample size of one and a brief phone call. A customer-related question really requires additional confirmation, but if during the customer discovery interview process you have a half a dozen people asking to buy your product before they know what it is or how much it costs, we consider that pretty strong validation. I can’t imagine what would have happened to me if I’d presented my 50 surveys and only 6 said they were interested. My guess is no degree would have been forthcoming. But we pass people with less now, routinely.
Why is such a small sample ok? Well, for one thing, the internet is a big place. I like showing my students an example of a very, very niche product, a chess set made from dead mice. When I first started sharing it, there was only one person selling it that I could find. Now there are many more, and a lot of variety. You can get dioramas of pivotal scenes in movies, only with mice. You can get mice in yoga poses. You can get mice in graduation robes to celebrate life’s big moments. It goes on.
Another thing that has happened over time is that we’ve got so stinking rich as a society that we have the free funds to spend money just to please ourselves. That’s pretty great for the purchaser who gets top buy something that makes them happy, at least for a moment. And it’s great for the entrepreneur who, up until this period of practically boundless prosperity, would have had no chance of making a living by stuffing mice and transforming them into Jules and Vincent, laying waste to a few Big Kahuna Burger eating teens in Pulp Fiction.
Also, identifying a target market, a beachhead market really, and finding out what they really want and are willing to pay for by experiencing deeply their pains and desires uncovers not just individual needs, but needs of the entire group. Even if the group is minuscule, according to Proctor & Gamble, it is quite often large enough to support a lifestyle business. In some cases it’s large enough to support something scalable while the entrepreneur finds the rest of the market and figures out how to engage with it.
This train of thought started with the concept of small data. I knew when I began thinking, and began writing that this was not my invention. It’s too obvious. We’ve been practicing it for a long time. I’d just never thought of it that way before. I was sure someone had, and thank you Wikipedia for letting me know that ‘small data’ is a thing, has a few competing definitions, and a lively conversation around it. Here’s one definition.
“Small data connects people with timely, meaningful insights (derived from big data and/or “local” sources), organized and packaged – often visually – to be accessible, understandable, and actionable for everyday tasks.” — Alan Bond
It appears Martin Lindstrom has written an entire book on it that is much closer to my meaning. Even he though, says get the small data to mine it for insight. Nascent entrepreneurs don’t mine though. If you’ll forgive the analogy they don’t have the equipment, the resources, or the expertise. What they do is synthesize the data they are able to get and then make generalizations about the larger world. It’s a very touchy-feely process. Heavy on thinking, feeling, and guts. Light on analysis, spreadsheets, and logic. But it works, at least for some. Entrepreneurship scholars criticize entrepreneurs for generalizing from a small sample, then in the classroom cite Jobs not doing customer research and Ford offering his automobile in ‘any color the customer wants, so long as it is black’. I think it’s time we accept the utility of small data, full stop.