What Is Attribution?


By Jay Adamsson

When analyzing your Google Analytics data, one of the first things you have to deal with is attribution.

What is attribution when it comes to website analytics? It’s about giving credit where credit is due.

Analytics, when done properly, is all about making decisions. And often those decisions are asking about digital marketing channels. For example,

  • Should I spend more on pay-per-click? Or should I spend less?
  • I spend a lot of time on social media. Should I scale back?
  • Was that banner ad I purchased on the newspaper website worth the money?

In all of these instances, the question is essentially asking whether an investment (either time or money) is worth it. And to answer that question, first you have to determine the value of the investment. In other words, what was the return that the investment brought to my business? And is that more or less than what I put into it?

Let’s take the simplest situation. Say you have an online store, and you can track the amount that everyone spends on your site. It would seem simple – take all the transactions and see where those visitors come from. So, if there are 20 purchases of $100 each, and 5 of them came from pay-per-click, then the value of pay-per-click is $500 in revenue. Correct?

Well, not quite.

The issue is that people do not only visit your site one time. A typical customer journey takes many visits. Maybe they first saw your site through a pay-per-click advertisement. They took a quick look at your site, but didn’t have time to look too closely. A couple of days later, they remember your site, and search for the your store name. They look around some more, and like what they see so they bookmark your site. Then later that day they click on the bookmark and look some more, going so far as to add items to the cart, but don’t finish the purchase. Finally, a day later, they come back to your site and finish the purchase.

That’s four visits to your site – the first through pay-per-click, the second through organic search, and the last two directly.

With the way we calculated value above, pay-per-click had no value. But this is in fact how they first found you. And that has value. So that $100 purchase should be credited not only to a direct visit, but the pay-per-click should get some of the credit. After all, that was how they first found your site. If it wasn’t for that investment, they would have gone elsewhere. And your calculation says that pay-per-click has no value.

Attribution is taking the value of a visitor, and spreading that credit over all the different ways that visitor came to your site.

Fortunately, Google Analytics has made it much easier to include attribution into your analysis. It is now a feature built right into Google Analytics. However, there is much more to it than simply clicking on a menu item.

Take our four-visit example above. Pay-per-click was the first of four visits. So should it get one quarter of the credit? Or should it get more because it was the first? Should the last visit get more than the other visits? If the four clicks happen within two days of each other, should they be treated the same as when the four clicks are spread out over two months?

As with most topics in Analytics, getting the data is only part of the issue. The more important factor is what you do with the data. How do you spread the credit out over all the visits? This is not a decision for Google Analytics. This is a business decision, and should be based on your particular business situation.

So, when analyzing your Analytics data, don’t forget about attribution. It’s important to give credit to not only what happened today, but also what happened in the past.