Attribution is Not Broken – It is Just Answering the Wrong Question

by Skip Fidura

The first direct response advertisement was written in 1477 by William Caxton in England.  He was trying out this new tactic to sell prayer books and hymnals. History has recorded that this was the first advertisement but has not recorded if it worked. The fact that we all have jobs in marketing leads me to suspect that it did.

Fast forward to the late 1800s and we find one of the most famous quotes in marketing and one of the most ironic quotes of all time.

‘Half of my marketing spend is wasted. I just don’t know which half.’

The sentiment is not ironic; it has been felt by every marketer since William Caxton. The irony is that this quote about attribution has been attributed to two people. If you come from America, you think it was uttered by John Wannamaker, a Philadelphia-based retailer. If you come from the UK, however, this was first said by Lord Leverhulme, founder of Lever Brothers. The reality is, however, that there is no evidence that either said it.

The fact is, business leaders need to know which investments are improving the bottom line. Those that are not paying off need to be stopped, while those that are helping drive performance need to be maximised.

Unfortunately, it is impossible to prove a negative. So instead of finding which half is wasted, we work very hard to prove which half is effective. Model complexity presents the first challenge with this approach. Simple models are easy to set up, use, and explain but in their simplicity, they miss out on many, if not most, of your touch points – think about first and last touch models that give all of the credit to one touch in the customer journey. Complex models will be a lot more accurate but are very hard to set up, are not user-friendly, and are very hard to explain.

The second challenge is that they cannot look at activity outside of the customer nurture journey. This could be your activity, like branding campaigns or the sales activity that happens outside of the funnel. Similarly, other external factors like the Brexit referendum or weather events are lost in your attribution models.

The biggest problem, however, is that attribution modelling does not answer the most fundamental marketing question: ‘If I had more budget, how should I spend it effectively?’

Imagine a senior leadership meeting where the CFO says they have found an additional £50,000 that can be spent this year. How would you answer the question? Think about that for a moment.

I can tell you how the CTO would answer that question. They would open up a spreadsheet, add in the £50K, and within seconds be able to say how they would spend it, detail the benefits for the business, define the payback period, and what the ROI would be on the spend. If the CEO is from an accounting background, then marketing would have lost that £50k as soon as the CTO pulled out their computer.

Our current approach to attribution modelling is broken because it asks the wrong question. Instead of asking which half of our data is not wasted we should be asking:

‘What marketing activity is driving the most value to my business?’

Every marketer since William Caxton has complained about not having enough budget. Until we can prove the value that we drive for the business, however, we will struggle to justify more budget, more resource, and better technology for our marketing.

To learn more about Skip check out his website.