Economies of Focus

How can small start ups compete with blue-chip giants? I've always wondered about that. Being somewhat of a (recovering) pessimist, it always seemed to me that the little guys have not chance. Yesterday, I watched this video:



The video is about a small company BitPusher that provides web operations and consulting services very similar to big players like IBM. The speaker isn't great, but the lessons are interesting:

• start out focused on a small set of services
  
• acquire and/or develop human expertise
• naturally migrate from one off to automated solutions
• one off -> copy & change -> template -> automate
• IMPORTANT:
  
• avoid temptation to automate things right away
• systematically keep track of all internal complaints about the current process
  
• prioritize: upgrade only parts of the process that really need upgrading

A one line summary would be: start out focused but have a vision (process) for growth.

This video echoed few thoughts:

Focus is crucial.
Small businesses should focus their efforts (only big businesses can afford to diversify). For me that means that I should try to implement the core of my idea. I often find myself coming up with ornaments that make the idea look more appealing. I guess I am counting on some sort of synergy between small ideas. This presentation is yet another reminder: stay focused.

Extract instead of inventing.
This idea isn't new either, but we make this mistake of putting the cart before the horse way too often. People in the computer science love to invent architectures and frameworks. However, only a tiny portion of these end up being used anywhere. Agile development and extreme programming books stress the same aspect: abstractions and generalizations should be driven by demand only. Which brings up to the next point.

Keep a log.
There is nothing groundbreaking here either. This guideline just says that one should learn from history. If some practice (or an idea) has been proven inconvenient (or inaccurate) note it. That way you'll have a list of things you can reconsider later.

Use your log.
Clearly keeping a log isn't going to solve any problems by itself. However, it helps you see the patterns. You'll know what you need to fix. The important thing here is to review your log regularly and to apply changes to your process.