Join Matt Bailey for an in-depth discussion in this video Numbers without context, part of Learning Web Analytics (2013).
One of the problems with analytics is the use of numbers in reporting and as indicators of success, without knowledge of the context that's necessary to give those numbers meaning. For example, we'll see this in a report many times, where visitors equal 200,000. I love to ask people well is 200,000, is that good, or is it bad? For some people they initially say it's good or bad, depending upon what kind of site you are. But eventually after thinking about it for a few seconds, people come to the realization that, well, that depends. And that's the answer.
You see, it depends. On what? Visitors equaling 200,000, it's not good, and it's not bad. What it is, is data. It's a piece of data, and data is value neutral. There is no good or bad value to data. And this is simply data. We don't know anything surrounding this. We don't know the time frame. We don't know what page. We don't know anything other than what we think it means. The same thing when we start looking at additional numbers. Page views equals 800,000.
It's not good or bad. It's a piece of data, because it doesn't tell us which page views. What's the time frame? Is this a day, an hour, a week, or a month? There's not enough data to ascribe value. So even when we talk about something like average time on site. What you're doing is taking thousands of visitors with thousands of different reasons, and putting them together and saying well, their average time on site was 6 minutes. This is value neutral. It's neither good nor bad, it's just a number that doesn't explain anything. And so, in order to move beyond our initial impression of what analytics are, we need to realize that our first reaction to analytics, which is looking at the numbers in the admin panel, and assuming that they mean something. We need to understand what they actually mean.
And that is, it's not good or bad, it's just a piece of data that doesn't help you improve your website. You see, we're talking about data, and data, it's not good or bad. Data needs context in order to be understood. And so, in your reports, when your reports are very simple, and contain not a lot of information, go back and look at your reports and see how much of your report is built on pieces of data. And because your report is built on data, a clear path of action is impossible, because you're not looking at the context that surrounds the data. And so in order to have a good analytics program, what you need to do is become more complicated.
And I'm sure that doesn't thrill you, but if you want a plan to improve your website, and improve your business, your data needs to be more complicated. So what do we need to do? Number one, have clear goals. What is it that you want to accomplish? And then those goals have to have context. If you know where you want to go, how are you going to get there? What does success look like? And then you need to build relationships. Beyond just the number of page views, or the number of visitors, how did they get there? What did they do? Build an understanding of what success will look like. And make sure that in your future reporting, you report according to your goals.
How did you meet those goals, and how did they enable the success of the business?
- Defining analytics terminology
- The problem of numbers
- Building segments for comparison
- Finding value in your marketing
- Creating valuable reports