How It Is

What Is The Measure of a Site?

What are you measuring? What is useful? What is the right way to explain it all?

After you think about where you’re saving your data, internally or externally, you’re going to be faced with the biggest problem known to exist.

What do you do with your data?

Common Data is (Mostly) Obvious Common Data is (Mostly) Obvious

Some data, as I’ve said before, is obvious. That is, you know what you want to do with statistics of visits. The base outset is ‘figure out how many people visit my site.’ Right? Not too hard. But that isn’t all you want to know. You want to know when your site is busiest, what content people read, and maybe you want to know on what device.

You want to know these things because they can help you optimize what you do next. If, for example, your Monday posts are super popular, then you want to make sure you post them at the time the most people are going to visit your site. If you know only 2 people view your site on an iPad, maybe fixing that little annoyance can wait a bit.

Top ↑

Rare Data is A Headache Rare Data is A Headache

On the other hand, when you look at statistics for your complex data, like a site with TV shows and characters and actors, you have a completely different problem. What public stats are both relevant and meaningful? And how do you represent them in ways that people can understand?

Like, do you use piecharts?

An example of two pie charts

They can be helpful but only if you don’t have a large number of data slices.

I made a pie chart with 28 slices and it was unreadable. Though that was mostly because everyone had between 1-5% except for one that had 75%.

Top ↑

The Question Is Usage The Question Is Usage

This is a problematic question because it has no easily defined answer before you start building out your site. We’ve all seen an image of a paved path and then a foot-trail cutting away from it, or winding around an obstacle. People like to joke about how it’s design vs usage. While our goal when making any product is to avoid people walking off the paths, it’s unavoidable. And in the case of public statistics, it’s even harder to predict usage.

A large reason for the problem is what is called a failure of imagination. This is, in part, the fault of the designers. That is, they didn’t predict things properly. Which requires metrics. Which can’t be gathered until people have used the site a little.

You see the problem, I hope.

Top ↑

Start With The Easy Start With The Easy

When I built out stats on my site, the ones I wanted people to use, I made sure to start with some easy things. Like those pie charts. Those are just pulled from a custom taxonomy which every character has. They’re simple. They’re easy. And they let people visualize.

After I released it, someone asked “Could we have a chart to show how many actors a character has?”

Actors per Character

That was actually not easy, but the point is that by starting with something ‘easy’ I was able to inspire people to ask what they wanted to see.

Top ↑

Don’t Be Afraid to Be Wrong Don’t Be Afraid to Be Wrong

Remember I mentioned that evil pie chart? You’re going to be wrong. You’re going to assume that the best way to show a specific data point is a pie chart when it really should be a bar chart. If you pick the right chart systems, it shouldn’t be too horrible to switch between them. But sometimes it will be.

Just remember, it’s okay to make mistakes. You can dig up a path and repave it after all.