Web Analytics and Attribution

Wikipedia defines attribution as a “journalistic practice of attributing information to its source.”  In analytics, it’s the struggle to understand which of your marketing efforts truly have a positive effect of your top and bottom lines.

It seems simple enough, you send an e-mail or run a PPC campaign and you get a sale.  Your web analytics platform tells you that you sold 7 widgets in one transaction and the visitor who made the purchase came from a Bing PPC campaign that was fired by the keyword “blue widgets”.   That sounds simple enough:  A click in Bing led to a $100 sale.  Not bad.

In reality, most sites do not have easy, one-visit conversions. Let’s dig deeper.  Using your web analytics tool, you are able to see that this new customer has come to your site five times over the last month.  With five touch points to the site; should Bing really get all the credit?

In Google Analytics, the last click that led to the sale gets full credit and this is known as last-click attribution.  If you don’t consider this truth, you are doomed to lower quality decisions regarding your marketing.

In Google analytics, there are a couple of options to help you get to more meaningful data:

1. You can modify the Google Analytics Tracking Code to, by default; give full credit to the first touch point with your site. (GATC is the snippet of code that is placed on every page of your site.)
2. You could use custom variables and unique landing pages for each marketing effort.

The first option may work if you are only interested in the first way someone became aware of your site and you have a short sales cycle.  If, on the other hand, you have a long sales cycle and many marketing channels, this method can be less effective.  For example:  If you use this method, someone that visited your site 6 months ago via Twitter and forgot that you existed;  but then  made a purchase after an organic search would have the value of the transaction attributed to Twitter and not your SEO efforts.

Using custom variables and unique landing pages allows you to “tag” visitors as they respond to your marketing.  It’s not perfect, but it can help you develop a better picture of what’s generating revenue for your business.  Let’s revisit our example from above, where on the fifth visit someone made a purchase via Bing PPC.  Remember that we attributed the sale to Bing.

If you used custom variables and unique landing pages, the picture could look something like this:

Visit 1 was from a display ad via the Google Display Network

Visit 2 was from Google PPC keyword “widgets”

Visit 3 was from your re-marketing campaign on the Google Display Network

Visit 4 was from Facebook

Visit 5 was the Bing PPC ad for “blue widgets”

So what you initially thought was a $100 sale on a $4 keyword in Bing has become a much more robust picture of how your multi-channel approach is paying off.  You can decide for yourself how much weight or credit each channel should receive.  Regardless,  you now have a much better understanding of how your marketing is performing and can make better decisions on which channels are deserving of your marketing spend.