Asking Questions Of Your Data

When asking a politician a question, you better ask the right one; otherwise you are likely to get answer, but it won’t be the answer to the one you thought you asked. When asking questions of your data, you also need to be careful how you ask it.

When most people start off using Google Analytics or some other analytics tool, they think amazing insights will just come to them through their tool. They slap the code on their site, wait a few days, login to look at their data, and see a roller coaster of a trend line go across the screen. If the roller coaster is going up, things are good. If not, things are bad.

But, if you really want to get insights from your data so you can make key business decisions you have to learn how to ask your data questions. So what types of questions should you ask? Below are some examples to start with and a recent client example.

Well you might start off with something as simple as how much traffic did I get? But, you realize without something to compare your traffic to, it doesn't mean anything. You then might ask, how many people filled out my lead gen form? Now we are getting somewhere, direct results. Then you might ask, what is my conversion rate, or in other words the percent of people that came to my site and filled out a lead. This is a great question. Now you might ask, how can I increase this?  So you ask, what traffic source has the highest conversion rate? Now you are hitting on the insights that are going to improve your website. A corollary question to this would be what types of customers convert more? This could be based on geography, technology used, products they looked at on the site, etc.

Criteria For Your Questions

So looking at these questions, what makes up a good question to ask of your data? I would say the following:

1. It has to be something that relates to the goals on your site.
2. It has to have some sort of comparison between other metrics on your site.
3. It has to be something that you can take action on.

Recently we had a question that we wanted to talk about with a client. We had noticed by looking at the data that their percent of mobile traffic had gone up considerably, and everyone is talking about the mobile web taking over. So our overall question was, “Should we build a mobile friendly website to improve the performance of the site?”

The question we needed to ask of our data based on this premise was, “Is there a difference in mobile and desktop traffic when it comes to conversion rate and revenue per conversion and does this difference justify the cost of a new mobile website?”

This question deals directly with the goal of our site, it compares two segments of the audience, and it is something we can take action on.

Putting Your Questions To Work

So how did we answer the question we created?

1. 1. We segmented out mobile & desktop traffic.
2. T  Then we tested to see if there was a statistically significant difference.
3. F  Finally we projected out estimated traffic over a year and the amount of money that is lost each year due to not having a mobile website.

This helped our client learn that they were losing around a quarter of the potential traffic because they didn’t have a mobile website. The decision then became a no-brainer because we asked the right question of our data.

Tim Flint is the owner of Flint Analytics and helps companies tell the story of their web analytics data so they can optimize the performance of the