Marketing attribution Q&A with Gabe Hughes
Written by Harry Groves
15th February, 2016
We spoke to digital marketing expert Gabe Hughes to get his view on the best way to set up marketing attribution where you work.
1. What does attribution mean?
Marketing attribution is the art and science of assigning sales or sales revenue to the marketing touchpoints which a customer was exposed to before to their purchase.
Although associated with the world of digital marketing, the problem of how marketing drives sales is as old as marketing itself. Attribution is therefore just a type of marketing effectiveness analysis, concerned with data rich, multi-touch marketing, and attribution today is concerned with both offline and online.
2. What advice would you give to marketers looking to adopt an attribution solution?
Two pieces of advice: (1) do not be tempted to ignore attribution; and (2) if you outsource the problem, stay close to the issues and do not simply take attribution analysis at face value.
In other words, you should have a handle on what the attribution issues are for your business. If you do not then your marketing measurement will be biased, and you will waste valuable money and effort which could be used to grow sales. Many firms believe that the problem is too difficult to engage with, and so postpone giving attention to the problem.
No attribution solution is perfect, and all providers of attribution solutions have reason to understate the limitations of their particular approach. If you are already engaged with the issue, then you can challenge the analysis and ask the hard questions. How is a solution addressing the attribution issues that are specific to your customers and your business? Be suspicious of any ‘black box’ solution that gives you results without explaining how your unique issues have been addressed.
It’s your marketing budget driving your sales – and so it should be your attribution which counts.
3. What are the most common mistakes that marketers make when it comes to attribution?
The first most basic mistake is the failure to realise that ‘last click wins’, is itself an attribution model. Companies who ponder whether they should ‘do’ attribution are in fact doing attribution already with correcting for measurement bias. Attribution is not optional.
As companies get closer to the data, another common mistake is to see quick single touch user journey data as evidence that the attribution problem is limited to a small group of customers. In fact we know that our multi-touch data is flawed; people online use multiple devices and browsers, and delete and block ad trackers. We also do not see the exposure to competitor ads. In other words, the raw data makes customer journeys seem shorter and more fragmented than they really are.
Finally do not make the mistake of letting the perfect be the enemy of the good. Attribution is a hard problem and no one approach is perfect. Attribution analytics should be about developing a better awareness of the sources of bias in your marketing measurement, and developing strategies to address to these.