Data Means Nothing

Data can mean anything, therefore it means nothing. How, then, can we extract meaning from a dataset?

Two shoe salesmen were sent into “darkest Africa” to feel out the potential  shoe market. The first telegraphed home saying: it’s hopeless stop nobody here wears shoes. The second telegraphed back saying: it’s wonderful stop nobody has any shoes.

As Product Managers, we work with a lot of data. Sometimes we even hire a data scientist to go through our data and tell us what it means.

Framing is the lens through which we view data. Since we all have different brains (and hence different frames) the same data will always represent something different to different analysts. Data can be objective. Recommendations based on that data can never be.

Data can mean anything, therefore it means nothing.

How, then, can we extract meaning from a dataset? One very effective method is to look hidden assumptions.

Suppose your analytics show a temporary downtick in traffic during February. One might assume that this is simply a natural ebb, another may assume that February must be a low month in your industry, a third may assume that there was a technical error has since been resolved.

To extract meaning from these analytics, ask yourself:

What assumption am I making,
That I’m not aware I’m making,
That gives me what I see?

Challenging this assumption will help you learn something new about your product (a competitor launched, an industry event, a political influence), which you can then leverage to your advantage.

Props to Rosamund Stone Zander and Benjamin Zander for inspiration.

Author: Luke Carbis

Luke is learning to trust his intuition, value Practice over Theory, and write.