An interesting and accessible book to understand predictive analytics is the book “Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or Die” from Eric Siegel.

What can we learn from predictive analytics while creating a startup?

Important for the success of a startup is in first place not to run out of resources before finding the right traction to earn enough to “live” or to get an investor. Traction is finding the right correlation and eventually causation.

Back to Predictive Analytics. The idea from Predictive Analytics is to find patterns in Big Data sources. Find correlation and if possible causation.

What does correlation means and does it always means a causation?

An example to illustrate correlation and causation from the book Predictive Analytics:

Increased ice cream sales correspondent with increased attacks. Why do you think that is? A casual explanation could be that eating ice cream makes taste us better to sharks:

shark-predictive-analytics

But another explanation is that, rather then one being caused by the other, they are both caused by the same thing. On cold days people eat less ice cream and also swim less; on warm days the opposite:

Warmweather-predictive-analytics

But how can we use this knowledge in Lean Startup?

In general with startup you need to find a correlation and if possible causation to generate users or better clients. In other words you need to find the engine of growth.

A good way to do this if you only have Small Data where it is not possible to find patterns or you simply don’t have the knowledge you can make test hypothesis from it.

For example, your startup create a test a problem-customer fit with one or more landingpages with a call-to-action to subscribe to your service.

What is predicted?
Which potential user will subscribe on your landingpage?

What is done about it?
Approach potential users (early adopters) in various predefined ways that are more likely to respond. Test if these potential users respond, try to segment and test the difference.

Next step is to collect more data and find the correlation and eventually a a causal explanation. This means your startup creates traction.

Leave a Reply

Your email address will not be published.