With the recent news that Uber is valued at $40bn (more than most of the Fortune 500) I believe the real question is – how big is the market?

The companies that have created the most value in recent history all have huge markets. For example, Google in online advertising and search, Facebook in social networking, LinkedIn for recruiting, Whatsapp in chat and Amazon in ecommerce. Having a large market allows a company to grow for years by focusing on what they already do well before needing to worry about new products or changes to their model.

The challenge with fast-changing technology industries and also with old industries (taxis) turned over by new applications of technology is working out how big those opportunities are. In my opinion, this all comes down to a clear definition of the problem being solved and in the case of Uber, this is the core of the debate about valuation. See for example NYU professor Aswath Damodaran “Uber isn’t worth $17bn” vs. Uber investor Bill Gurley’s “Alternative Look at Uber’s Potential Market Size“. In essence, the $40bn Uber question is whether it is a better way to hail and pay for taxis or an alternative to public transport and car ownership.

It is easy to see how Uber makes taxis easier, you don’t have to stand in the rain or cold with your arm out, you don’t need cash and you don’t have to carry paper receipts. I don’t think there is any doubt that taxi apps will consequently dominate the taxi market over time. However, it is incredibly hard to quantify Uber’s impact on public transport or private car ownership based on historical data as it hasn’t existed before! For investors, being able to judge markets that don’t exist yet has led to successes like Amazon, Google, Facebook, Twitter and Shazam.

In my own case, Uber is one of the reasons I no longer own a car and I have consciously used it instead of public transport. However, until so many consumers give up cars because of Uber that they partner with used car dealers, I struggle to see $40bn of value – especially when you consider switching costs.