09 Dec 08 What business problem are we solving?

“Technology in search of a problem,” is a longstanding Silicon Valley standard criticism of companies with smart people who build cool stuff but fail to generate revenue. Sometimes it seems like the entire world of web analytics could be described that way. The squishier the metrics definitions are (prime example: “engagement”), the more accurate the description is.

The most important word in that description is “a.” In other words, find just one problem to solve at a time, rather than a dozen. This is a normal problem in a developing environment, where inventors are driving. The problem is that people who are good at inventing products and services are also good at seeing problems they can solve, so they do a lot of things moderately well and excel at none. In the day-to-day work of web analytics, this often appears as “data smog,” a term I first encountered in Actionable Web Analytics, which credits it to David Schenk.

Somebody involved in every analytics effort absolutely needs to be talking to the ultimate customer, the one who is actually generating revenue, to understand their business. At the very least, analytics should address the specific business problem a customer knows it has. Ideally, it goes further and sees solutions or opportunities that the customer didn’t realize that data could address.

In other words, I think it is a mistake to just ask or settle for being told what numbers to deliver. Ask why the numbers are needed. Ask for the goals and their priorities so that you can set analysis priorities. For example, the goal is support, the speed of responses in the social network matters a lot more than when the goal is loyalty.

If there is no way around the need for a lot of measurements, fight data smog by knowing your priorities and packaging the data with some sort of drill-down that puts the most important numbers on top. This is where the dashboard idea becomes critical – hide the complexity behind a one-page (or less) summary. The fact is that even when people think they want complexity, they almost never do. I created a set of Excel-based dashboards with deep drill-down… and I’ll bet that hardly anyone used more than a small percentage of what was there. But it was there if they wanted it and that keeps people happy.

All of this adds up to one word: focus. A venture capitalist friend gave me a mantra that stuck: the five most important things (for startups, but it applies to innovation in general) are focus, focus, focus, distribution and focus.

Tags: analytics, focus