One of the best things about online services is the data. The amount of data can get massive quickly.
We all know that data is valuable for a variety of reasons.
But the thing I want to talk about is how that data can make your product better (vs data monetization).
Many web services do capture and record vast amounts of data. But a lot of that data isn’t very helpful. The key is identifying the important parts that can improve your service.
First, if you aren’t capturing that data then get busy and start. I recently met an entrepreneur who created a widely used Firefox extension. But he wasn’t keeping track of any data. He is now :)
Next, find out what data is important. Well, that’s the hard part.
Few examples to consider.
I recently spoke to an executive that ran a successful subscription based virtual word. He told me that as soon as users invested x hours into the product they were subscribers for life. The assignment in their case was pretty clear: focus on user engagement and the business model will compile. Other data was helpful but much less important than getting users to invest time in the game.
Another entrepreneur told me that in his social app, if a user connected with 5 friends then there was a high likelihood they would become highly active. They are working on improving their invite a friend component and friend discovery. Other features are taking a back seat for now.
Yet another entrepreneur told me that reducing the sign up process to a bare minimum, user name and password and nothing else, generated a 3x increase in sign ups. They did a lot of A/B testing of their home page & sign up experience.
In all of these companies the critical data components were different from one another.
A lot of this seems like common sense but it’s a challenge to figure out and prioritize. That’s especially true in a startup where you are chronically under funded and understaffed.
But I encourage you to make the investment & track down the good data and ignore the useless stuff.
(n.b. the entrepreneurs mentioned in this post are not in the Spark portfolio).