Lies, Damn Lies, & Attribution Analytics
“Half of the money I spend on advertising is wasted. The trouble is I don’t know which half.” – John Wannamaker
Over the past seven years, an avalanche of advertising dollars has been redirected from traditional media channels to digital or online media. One of the driving forces behind this trend is the hope of stronger, more accurate attribution – the ability for marketing professionals to more easily and accurately assess a true ROI for their precious advertising dollars. The ability to track, measure, and retarget based on who has seen and interacted with an online ad is great – especially if the ads lead to a purchase that happens entirely online, such as with a flight or clothing purchase. However, many businesses still require customers to interact with them in the real world. Attempting to tease out the secret of attribution can be complex at best and unattainable at worst.
From a 30,000 ft vantage point attribution analytics attempt to show the steps in a buyers journey and allocate credit to those that participated causally in the conversion. Unfortunately, there are usually a few steps a buyer takes that don’t contribute at all to the conversion yet they receive undue credit for being part of the process. Additionally, there can be unseen steps taken that never get accounted for in the attribution analytics models.
Challenge #1- Wasted Budgets
More than 50% of digital ads are never seen. comScore reports that this number jumps to over 69% for network/exchange/ programmatic display networks. These statistics suggest that a huge portion of a marketing budget gets wasted. The real problem with ads not being seen is trying to track which ads were not seen and to what degree. This leads to the next problem.
Challenge #2 – False Positives
One of the promises of Attribution Analytics is an accurate revelation of what is working and what is not. Imagine you are looking at your Analytics report at the end of the month and you read that impression Ad1 generated 100 leads and impression Ad2 generated 20 leads. The natural assumption is that Ad1 is outperforming Ad2. The report suggests that Ad1 works and Ad2 does not. Money from Ad2 is redirected to Ad1. In reality it can be difficult to impossible to know if each ad was seen in equitable volumes. The result is we fire or discard ads we perceive are not performing as well when there is often very little information to support such decisions.
Challenge #3 Offline Media
Despite the fact that offline media works incorporating it’s impact into an integrated conversion funnel is tough. In a survey report released by the CMO Council nearly half of marketers said integrating offline media and physical customer experiences into a comprehensive attribution model is “selective at best.” Most new technologies and platforms attempting to address offline/online still back into attribution results rather than to provide a clear line between an action and a conversion.
Challenge #4 Cross-device attribution
Another difficult gap to bridge in measuring steps in conversion is tracking consumer behaviors and interactions across the multiple devices used today. Buyers are very comfortable watching an ad on TV, jumping on a computer to do research, switching to a phone to shop, and ultimately purchasing in a store. As buyers migrate between devices it become almost impossible to follow their pathways through unique IDs assigned to each device. The attribution process fragments and conversion credit is ultimately assigned incorrectly. This process will become more murky as search migrates from keyboards to voice-commanded devices like Google Home and Alexa.
In Conclusion..
The goal of good attribution and analytics is to be less wrong today than yesterday. A general understanding of your buyer’s journey is better than nothing, but the specificity and granularity needed when allocating marketing resources demands that even the squishy areas of attribution and analytics get accounted for and properly addressed.
