8 August 2011 | Team Tamar

Finding your way through the attribution maze!

There’s a lot of talk online about attribution modelling, the why’s, the wherefores and whether online brands should even bother invest in one. You’ve almost certainly given it some thought over the last year. However, even though you may not think you are not currently using an attribution model for your online brand, you are in fact probably already using one.

Surprised? Don’t be. If you’re using Google Analytics, Yahoo Analytics or a similar analytics tool, that is actually an attribution model. It’s a ‘last click’ last attribution model. You no doubt already know its limitations. You’ve probably been pulling your hair out trying to find the answers to your pressing questions: “Why is revenue not as high as I expected?” “Why is traffic so low when I’ve spent so much money investing in it?” It is at this point that you should take a breather, and think about proper attribution modelling – especially if you find yourself currently struggling with the limitations of that ‘last click’ attribution model.

With one of Tamar’s clients I recently saw a significant and puzzling increase in direct traffic to their site. However when I spent some time analyzing the data, I realized that the tool they were using is based on ‘last click’ attribution. So, even though the end access point to the site is credited in the tool as the direct channel, there is a more than average chance that many other channels have assisted this conversion.  The frustrating issue with this model is that we don’t know which channel has contributed to this uplift.

Let’s take an example of a football team. I’ll use Man United (my favourite team). Rooney might have scored a goal in a match but he couldn’t possibly have done it without a pass or passes from other team players. Even though Rooney takes the ‘credit’ for scoring a goal himself we can’t forget all the team players (or referrals, in the online model) who assisted along the way in scoring that goal. A proper passing strategy and the efforts of all the team players have a huge impact on the desired end result – a Rooney goal.

It’s just as important to have a properly planned attribution model that makes the most sense for your revenue cycle and campaign strategy.  It is also essential to stick with your choice for the optimal analysis of performance across all your applicable channels.

To get started you will need to ask yourself “once I have my attribution tool in place, how do I design an attribution model?” Firstly, it is important that you understand that there is no definitive ‘right’ answer. There is a multitude of modelling approaches available. Below are just a few examples of the different types of models:

  • Even Weighted Across all Channels Model
  • First Contributor Attribution Model
  • Last Contributor Attribution Model
  • Bath Tub Attribution Model
  • Binary Attribution Model
  • Applied Attribution Model
  • Best fit Attribution Model

The list is big, but these are only a sample of the very many available models we have worked with recently at Tamar.

So, while the choice of attribution tool or platform is important, it is not nearly as important as finding the right attribution model that best suits the needs of your business. Whatever you do, don’t be afraid to try out a few different models until you find the right one – good luck!

Team Tamar