Monday, 5 August 2013

A lean recipe and market size assumptions

Making assumptions about the size of markets is important. It's why the big analyst firms like IDC, Forrester and Gartner have revenues that are often bigger than entire sectors that they produce reports on. (Do any of them report on the market for business analysis, I wonder?)

What's great for small businesses and start-ups is that much of this information is made public and is useful for understanding the market you operate in or would like to move in to. How is the market likely to develop over time? Core data about market size and forecast growth can help to validate key start-up assumptions.

What does your forecast growth look like in comparison to the growth (or decline) of the sector as a whole? Does that share of the market look reasonable at each key stage of the business? If you're looking at owning more than 100% of the market at any point then you clearly have a problem with your key assumptions.

But understanding market share on your growth journey can be important for a host of other, more strategic, reasons. What are your competitors likely to do as you acquire significant share? At what point do regulatory concerns have to be taken into account? What might consolidation look like in this sector?

Taking the size of a market should not, however, be used as the sole underpinning of a business plan and the assumptions about sales growth. This flawed argument works like this: We think we can achieve X in sales in year one, but we don't have any real basis for this assumption. It's a hunch. But X is only 0.1% of the total market therefore X is a perfectly valid assumption and we should therefore proceed and make decisions about the business on the basis of this forecast.

Imagine me estimating how many people I think will read this post. There are about 50 billion pageviews per day globally*. I'm pretty sure I can get just 0.002% of today's pageviews across the planet. That seems like a tiny proportion. Therefore, I'm going to say 1 million views is easily achievable and I'm going to pay for enough server capacity and bandwidth from my hosting company to deal with that demand.

The absurdity of this example is quite plain. Of course a more reliable guess would be to look at pageviews of previous posts. If none was available, maybe look at how many followers I have on the social networks that I post the link to and make an assumption about a 'conversion' rate from those channels. But in the absence of any metrics that could be used to even loosely estimate core assumptions, what do you do?

Well, you don't make any critical spending decisions or commitments on the basis of your initial assumptions. You have to take the "Lean Startup" approach. Build a minimum viable product and test it out. You can't start with a whole market and work back using a reductionist approach until you get to a share that sounds about right. That would be a sure-fire recipe for failure.

* source = top of head