In my last blog, I talked about the importance and value of visitors who return to your website having not engaged with you on their previous visit.
Having those people return is generally a good thing, since they’re more likely to buy or engage with you than even existing customers. And that means you want to know, as far as possible, the journey they took that led to them returning to you and which interactions played how significant a role in their decision to return and even buy. This takes us into the complex world of attribution modelling. Basically this is correctly identifying the traffic source responsible for every conversion your website receives.
Now you might be thinking that the ad that was most effective was the last one a visitor clicked on to return to your site. But that ignores the important role of all the other clicks this person made before that. For instance their journey could start with a paid click (from a product based campaign), followed by an organic search click, a social media remarketing click, with this torrent of clicking ending with another paid click (now from a brand name campaign) before they visit and buy.
The default attribution model Google AdWords applies places the conversion in the bucket of traffic responsible for the last click. That being the brand name AdWords click in the last example.
Obviously this causes inaccuracies when people are bouncing in and out of your website through various streams of traffic before they convert. By analysing conversions through the lens of Last Click those responsible for kicking off the buying behaviour are naturally de-optimised over and above those which close the sale. And you can’t close what you don’t start so revenue slides – downwards.
Thankfully Google AdWords offers several more attribution models to pick from. To compare to Last Click there’s First Click attribution. Here there’s total emphasis on what starts the journey. It’s another extreme weighting but now at the other end of the journey. Which is why the next four are where the money is.
Linear attribution shares the credit for the conversion equally across all clicks. Whereas Time Decay gives more credit the closer a clicks occurred to the moment of conversion. Then Position-based gives 40% of credit to the first- and last-clicked ads and corresponding keyword, with the remaining 20% spread out across the other clicks, while Data-Driven (the geek option) distributes credit based on past data for this conversion action (assuming enough data is available).
Remember this discussion is a non-discussion if your traffic behaves and arrives and converts all within the same session. In that case, you can use Last Click Attribution and feel comfortable living in the land of the default setting. For the rest of us it pays to pick a model away from the two extremes to see how it alters your conversion reporting. Google AdWords allows you to compare a couple of models at a time. More often than not, the most suitable attribution model for those with bouncing /cross channel traffic will be Position Based.
So to recap. Let’s say a customer finds your site by clicking one of your AdWords ads. She returns a week later by clicking over from a social network. That same day, she comes back a third time via one of your email campaigns, and a few hours later, she returns again directly and makes a purchase. The Position Based attribution model may assign 40% credit to each of the Paid Search and Direct channels, and 10% to each of Social Network and Email channels.
Two last points. First, with greater insight comes greater precision. So if you do make the switch out of the extreme choices then get used to measuring conversions to at least one decimal place (24.8 vs 24 or 25).
Second, this can be as mind bending as it sounds. While the effort’s well worth it, you should definitely call us for a reliable guide through this difficult, yet highly rewarding, jungle to find the attribution path best suited for you and your website.