Andrew Chen over at Futuristic play, wrote a great post on the importance of analytics, but also the balancing act of deciding what to track and how much time to spend on it.

With our portfolio companies, and startups in general, metrics are key to measuring your business performance and help to guide the company.

Andrew reminded me of the following key points:

• Some reports are operational. Some are investigative. But these are two distinct outlets.
• Be mindful of the engineering resources and opportunity cost of developing analytics tools and reports.
• Also be mindful of the valuable output these reports produce, which may prevent wasted development / false starts.
• Most importantly,If you’re not going to do something about it, it may not be worth measuring.

I’m guilty of overlooking these points from time to time, but will be more mindful going forward.

Have a read: omg I’m just a startup, I can’t do those fancy analytics!

The Angelsoft blog referred to our ongoing advice not to write a business plan. It generated quite a debate on that site - I encourage you to read through the comments if you are interested in the topic. They also shared a link to a one page pdf executive summary that they believe is very useful in fund raising.

Building a startup is like conducting a large scale experiment. There are many unknowns and lots of variables. Every step of a new venture has to be figured out and conquered as it comes. It is for this very reason that so many early stage companies perish soon after they are started.

However, after building my own company and being involved in countless others, I am pretty certain that there are frameworks that can be applied to most startups that will help reduce some of the inherent risk.

If you take the experiment analogy I used above, you can start to think of a startup as an exercise in mastering the unknown. Instead of focusing on what you think you know, which is how most entrepreneurs approach their business, concentrate on what you don’t know. Find the key issues that you need to understand and scrap out all of the variables around them. Then test your hypothesis with as few resources as possible. If you are diligent you will be able to eliminate a lot of the inherent risk of a startup and make better decisions.

Let me give you a theoretical example that I think will illuminate the point I am trying to make. Suppose you were trying to introduce a revolutionary new consumer oriented venture. Instead of taking a wild guess as to which marketing mediums will be most effective, you could set-up limited trial runs with the most promising options. Perhaps you would dedicate one thousand dollars to Adwords, postcards, flyers, and tradeshow exhibits each, and track the results diligently. In little time and with few resources spent you would have a pretty good idea of what works and what doesn’t, only then does a large scale expenditure make sense. This might sound obvious and intuitive, but you would be surprised as to how many times I have seen entrepreneurs (myself included) make quick intuitive assumptions and pour tremendous resources behind those hunches. Another quick example is the building of a website, most entrepreneurs get fixated on getting their site perfect before they launch and spend tens of thousands of dollars doing so. Instead, I propose that they get a quick and dirty site up that lets them get real customer feedback as to their idea. Once they have an idea of what their customers really want and do, they can more wisely spend their resources building out the perfect site.

Spend as few resources as possible proving out as much as you can, because your resources are much more valuable when applied in the right places.