Building a business is hard, running one while you are building it is harder. No wonder there are all these metrics on the failure rate of startups. Try juggling 5 balls while you are on a mono-cycle on the freeway, that is what it feels like when you are in the early stage of a startup. So, how does one navigate this mine field? Well, if we go back to basics, we absolutely need to have defined metrics that matter to building the business. Eric Ries in his book Lean Startup talks a lot about this but none of the stuff that he his talking about is new, just like what Warren Buffet says he does in the investment process is simple, but doing what he does is very hard. I want to make one small change to Eric’s method of Build-Measure-Learn, I want to change the order a bit Measure-Build-Learn. This idea goes back to the times when I was in software development, I actually got pretty good at it when I started writing Test Code before I actually wrote the Code for the Use Case, I saw that I was making rapid progress in how well I could write the actual functionality in the system. For all those people writing software, read the RefactoringTest Driven Development method and Continuous Integration by Martin FowlerKent Beck and others. The Lean Startup book refers to this as well.

Lets define some metrics that matter, this also depends on which stage of the development the startup is:

Idea: You have a brilliant idea that you have shared with your friends and family and they think that this could be the next best thing to sliced bread. The things that you absolutely need to instrument at this stage are:
  • Is your potential customer really hurting so bad that they are willing to pay you to solve it? how do you know?
  • How will you find your customer? or more importantly how will they find your solution?
  • What is the size of the market? how do you know?
  • Is the market growing? how do you know?
  • Will the customer continue to buy from you? how do you know?
The important thing in instrumenting your idea is to automate it, i.e just like a dashboard in your car build some gauges. The age of the Internet has bought all this to your doorsteps. There are many resources that are available that provide you with a wealth of data that you can read and make an instrument out of.
Startup: You have validated the above 4 things and you have also build a Minimum Viable Product (MVP), you have your first customer and you are excited that this thing is actually going to make you afford a cafe latte. Great! so what are the metrics that you absolutely need to have here?
  • All of the above +
  • What feature of the solution that you have sold to the customer that the customer absolutely cannot do without? how do you know?
  • How do you plan to scale your business? i.e. how are you going to sell more stuff? how are you going to support the customers?
  • How do you collect your payment? have you automated it?
  • What is your Average Revenue Per User? (ARPU)
  • What is your turn around time for customer issues?
  • How many “features” read “bugs” that need to be addressed right away?
  • How many people visit your website? (although the above ideas can be for any type of business, this metric is critical, if you are not being found you are not going to scale the business… Period)
  • What is your conversion rate? i.e % of those visiting your site want to try your product/service
  • What in your website attracts the visitors? this is nothing but which links they click on.
  • How is your bookkeeping? are you accounts in order? have you automated it? again no excuse for not using tools that are available today to automate this important metric.
Once again measuring all of this is trivial in the age of internet. There are many tools available to do that Google Analytics, KISSmetrics, etc so no excuses for not having built instruments into your solution.