Statistics, Metrics, and Analytics

So, I was recently presented with a set of analytics from a rather self-satisfied member of my team.  The analytics showed:

1. The number of views of our product,
2. The number of saves of our product, and
3. The number of times the product appeared in a search.

They were just raw counts gathered by a computer system.

In other words, analytics.

But they were presented as metrics, which kind of annoyed me.

Now, there's a big difference and the words aren't interchangeable like the current office usage.

A statistic is a stroke tally of something your measuring.  It could be the number of visits to your product, it could be the number of times a batter hits a ball- you get the idea.

An analytic is when the computer program is measuring the counts for you.

A metric is a normalized statistic, most commonly for measuring some business goal.

You know, normal...

...like the bell curve??

The difference here is that a statistic or analytic is a raw count.  The metric is some computation that you do with the statistic...like "the average number of visits per day to Product X" or "the difference in the number of views of the leading product per day compared to Product X".


Now, I'm not a statistician, but honestly, the first things you should ask yourself when presented with a page of numbers is:

1. What is normal?
2. How did you measure this?
3. What was your measured population?
4. Why do I care?

If you can't answer all of these by looking at the page, then you have a bunch of statistics- NOT METRICS.

Step 2:
Go back and do it right.

Step 3:
Stop using statistics, analytics, and metrics interchangeably.  You look like a moron when you do.