It’s common for VCs to look at the market size for a potential investment from a top down and bottom up perspective. The top down perspective takes market research, often from an analyst firm or investment bank and the bottom up approach works by multiplying the number of customers by their likely spend – more detail in my old blog post here.
What I hadn’t thought of until recently is that it’s also helpful to take a top down and bottom up approach to assessing likely demand for a product.
The top down approach looks at how a startup fits with prevailing big picture trends. At the time of writing AI is the trend of the moment and it’s a good starting point to think that companies which intelligently apply AI techniques can create useful products. Moreover, it’s also true that raising money is easier for companies that are on trend (investors love a herd… or at least most of them do!).
However, the top down approach isn’t sufficient on it’s own. Even though it sometimes seems like companies doing AI for XYZ seem to be raising money almost as easily as companies doing Uber for ABC were a couple of years back, this strategy is unlikely to yield much success for either founders or investors.
To make good investments it’s important to combine the top down approach with a bottom up approach which looks at use cases. If it’s difficult to convincingly explain how someone will use a company’s product, it’s a fair bet that they will find it difficult to get customers. I’m consistently surprised how often entrepreneurs allow themselves to be satisfied with only a vague understanding of why they will make people excited.
When looking from the bottom up, a good first question to ask is ‘what behaviour potential customers are already exhibiting which shows that they will have demand?’ For young software companies a classic answer to this questions is that potential customers are building homegrown versions of the product they intend to build. If our young software company can build a software product that’s better and cheaper than the homegrown version then it’s a fair bet these companies will stop writing their own code and become paying customers.
A second technique is to employ Clayton Christensen’s ‘jobs to be done’ framework which starts from the insight that customers buy things because they have jobs they want to get done. Jobs can vary from the mundane (e.g. cutting the grass) to the exotic (e.g. become my better self) and companies that can articulate a good fit with a job that lots of us have to do or want to do are in with a good shout of selling lots of product. There’s more detail on the jobs to be done framework here.
For infrastructure companies the use cases are often not end user use cases. Rather the use cases are to help other companies build use case for the ultimate end user. For example a company that makes electric motors might sell to a lawnmower manufacturer who’s job to be done is to sell more lawnmowers. The electric motor opportunity can then be evaluated on the basis of whether it will allow the lawnmower manufacturer to help its customers (the end user) with their job of cutting the grass.
As with market size analysis the bottom up approach is harder to do well, but yields much richer insight.