I was just browsing the Buffet FAQ – a compendium of Q&A’s with Warren Buffet – and my eye chanced on a question about buying companies where the numbers told him not to. I think this is very interesting, here’s a quote:
But if you don’t know enough to know about the business instantly, you won’t know enough in a month or in two months. You have to have sort of the background of understanding and knowing what you do or don’t understand. That is the key. It is defining your circle of competence.
Everybody has got a different circle of competence. The important thing is not how big the circle is; the important thing is staying inside the circle. And if that circle only has 30 companies in it out of 1000s on the big board, as long as you know which 30 they are, you will be OK. And you should know those businesses well enough so you don’t need to read lots of work. Now I did a lot of work in the earlier years just getting familiar with businesses and the way I would do that is use what Phil Fisher would call, the “Scuttlebutt Approach.” I would go out and talk to customers, suppliers, and maybe ex-employees in some cases. Everybody. Everytime I was interested in an industry, say it was coal, I would go around and see every coal company. I would ask every CEO, “If you could only buy stock in one coal company that was not your own, which one would it be and why? You piece those things together, you learn about the business after awhile.
I love this notion of a circle of competence and I think it’s very relevant to working in technology today, the more so as technology continues to fragment into more and more niche disciplines. A few years ago most competent developers could be on top of, say, deploying a web application. Now it’s Baas? Or PaaS? or maybe IaaS? Use a cloud? Whose cloud? Public or Private? Do we containerise? Which container? How to automate? Use Chef, Puppet, or Ansible? And this is just deployment!
Choice of language, tools, or frameworks have an even greater range of complicating factors. Now imagine the topic is blockchain or maybe artificial intelligence, you have to imagine your circle of competence shrinking rapidly!
Being a technologist today is about understanding your circle of competence, it’s boundaries, and knowing people with useful overlapping circles and how to work with them.
Further it seems to me that this notion applies double in the investment world. When investing in unlisted companies with unproven technology and/or business model how do you figure out if it makes sense? Even worse, good startups tend to be solving hard problems which, by definition, may be tricky to understand or very niche. As an investor how do you determine your circle of competence? How do you decide which of the 30 companies out of the 1000’s in your deal flow that you can get a good read on?