The three categories of start-ups

When people are discussing start-up companies there is often a disconnect around what is meant by a start-up.

In my view there are three categories of start-ups and discussing more than one of these without formal identification can lead to cross purposes.

The three broad categories are:

1. The disintermediation plays

2. The tech start-up

3. The science based start-ups

Other companies are started, eg franchises and cafes, but they aren’t startups in my books.

The disintermediation play has a single purpose – to cut out existing service providers in any distribution channel. It could be Uber with it’s $6b of investment in the taxi business, or your local self-funded pizza delivery start-up app. Using off-the-shelf and recently available IT technology, the development of the services may be completely outsourced. Most of the investment in these efforts go into sales and marketing in order to grab market share.

The tech start-up takes on real technology risk and can fail because the technology doesn’t work (to spec or at cost, or at all). In the modern era the tech start-up can sometimes look like a disintermediation play if it is in the web services/app space but is differentiated by, say, some big data algorithm development that can stuff up and provide no benefit. Alternatively the tech could be some new widget that takes a punt on replacing some existing consumer or B2B hardware. Generally, as soon as there is hardware there is considerable risk due to working capital issues and the hardware not working.

Science based start-ups are quite rare these days. Often based on university findings these ‘technologies’ can take years to find a market sector that cares (i.e. is often a technology looking for a problem to solve). Even when a problem has been identified, investment in such a start-up is ill-advised because the long time frame of getting the product to market and the risks associated with failure; these factors generally blow the IRR out of the water. Examples would be pharma (which is the most targeted of this category), and any of the former fiber-optic or laser start-ups of generations back.

The boundaries between the three types of start-ups can get blurred at times. Some start-ups pivot out of one category and into another when the going gets tough.

Investors usually stick to one category but sometimes spread across two in order to spread their risks.

Disintermediation plays are usually the cheapest to start. They have the lowest tech risk but also the highest market risk due to the massive global competition in each category. The winners in this space get good early traction due to great execution and then win by getting millions or even billions in investment to shove their capability down the throats of their customers. Incubators and accelerators are dominated by this sort of start-up. I suspect that technology in this space is evolving so quickly that it takes about 5 years, on average, for a technology solution to become obsolete and then the disintermediator is at risk of being disintermediated if they don’t continue to invest in updated technology solutions. My personal view is that the best founders of disintermediation plays have a deep experience and understanding of the market they are operating in.

Tech plays need good licks of early funding to get across the alpha prototype line. But they also have less competition as a rule and often need less investment than disintermediation plays to tackle the global markets they are taking on. Outside of silicon valley these deals have a difficult time getting funded at the moment because most of the risk capital is flowing to the disintermediation plays due to their apparent higher returns.

Science based start-ups are quite rare these days and often float along under the radar. Usually they run on the smell of an oily rag for years while they try and prove that the science can underpin some useful product or service that a customer cares about. These start-ups are the most likely to have meaningful patents. Investors in these types of start-ups are either bored, deluded or eccentric.

The purpose of any start-up is to rapidly develop the tech solution, create fast adoption and then also enterprise value, which needs to be crystallised through an IPO or a trade sale. The good start-ups work backwards from the end-point – that is, they have a vision of when they will list or who will buy them. And then they create a plan to execute to that point in the future.

Accepting this proposition, you will see why disintermediation plays are so attractive to investors compared to the other sort of start-ups. The only special skills needed are industry know-how, great sales and marketing, lots of capital and good luck. The other types of start-ups require additional skills, such as great management of the development of technology solutions, the risks associated with the technology not working, and the market not caring.