Multi-Attribution Versus Last Click Attribution Modeling


Most visitors making a purchase from a website have multiple marketing-driven interactions with that site prior to the final visit during which they make their purchase. Let’s examine a sample series of referrals to a hypothetical shoe retailer: In this case, assume we have a customer that hits our hypothetical website as a result of five different marketing events (in sequential order):

1. Google AdWords search for keyword “shoes”;
2. Bing organic search for “slingback shoes”;
3. Google AdWords search for “Manolo Blahnik Slingbacks”;
4. Referral from ad purchased on a shoe-related blog; and finally
5. Google AdWords search for our ecommerce website name.

Popular marketing analytics packages focus on the last marketing click made by the customer prior to checkout. In our example case, that would be a search on Google for the name of the ecommerce website with a click on the AdWords ad for the website. What that means in practical terms is that the person in charge of purchasing advertising at our company would get the misguided impression that she should focus spending on this phrase while eliminating spending on the keywords that actually introduced the customer to the website. In addition, she would think there was no value in the ad on the blog because it would appear that no referrals from this ad generated any sales, when in fact it may have been responsible for reassuring our customer that our website was indeed a legitimate source for these highly-coveted shoes clickfunnels pricing 2020.

We have dubbed this misguided approach of focusing on the last click the “last click fallacy.” The last click fallacy results in overspending on navigational ad terms while under-spending or eliminating marketing spend on the so-called “top of the funnel” terms that actually introduce customers to the website. Multi-attribution methods, on the other hand, take into account all of the marketing events that contributed to a conversion, allowing the marketer better visibility and more effective ad optimization. Furthermore, multi-attribution models are flexible; for example, a marketer can choose to over-allocate credit to the first or last marketing click rather than applying credit evenly across all clicks on a pro-rata basis. In our experience, the multi-attribution approach leads to the inclusion of terms that that deliver more customers at the same (or sometimes even better) ROI, but in order to achieve maximum effectiveness, it is necessary to capture information on all marketing events in a way that cookie-based tracking frequently cannot handle.

Leave a Reply

Your email address will not be published. Required fields are marked *