By David Herscott
Attribution Management is the latest attempt by marketing technology companies to provide advertisers and their agencies with a solution for tracking and valuing the slew of advertisements that lead up to an Ecommerce sale. Attribution Management has a more complete definition, but beware: this site is a ClearSaleing blog, one of the early technology leaders in the space.
Currently the default methodology is “last click,” meaning that regardless of the marketing messages that a consumer is exposed to, the last click – whether it be search, display, affiliate, email or other – gets the credit for the sale. In some ways not much has changed since the advent of modern advertising – brand marketers are still fighting it out with direct marketers to get the kudos (and the lion’s share of the budget). However, thanks to technology we can now create an entire picture of brand interaction prior to sale – well, sort of.
The technology has come a long way in a relatively short period of time, and attribution management can be an effective tool for Ecommerce marketers on the hook to show ROAS/ROI. However, the current challenge has less to do with technology and more to do with developing industry-accepted methodologies.
If you look at the early leaders – ClearSaleing (now a GSI/ebay company), C3 Metrics, Convertro and Adometry (you can even check out Visual IQ and Atlas) – you’ll notice that they all use similar attribution models. These models loosely include last click, equal attribution (eg 25/25/25/25); and a weighted-model that looks something like this: 60% credit to first exposure, 30% split among the second, third and/or fourth exposures and 10% to the last click. These companies also offer “custom models” which are algorithms based on prior results.
Attribution Management companies often use a weighted model to track a buyer’s steps to make a purchase.
Whats wrong with the current models? Last click provides no real value – any basic web-analytics solution will give you that. Equal attribution is not based on anything other than indecision. The weighted model is interesting but I have yet to see any compelling consumer-behavior data that supports this model. The custom model obviously has the most promise but I think that technology providers need to be more transparent about how the model works so that we, as marketers, can make better decisions.
David Herscott is a Managing Partner at NetX. Contact David at David@netx.com.
