The marketing industry is approaching a crossroads. With issues of user privacy gaining traction and priority in industry discourse, it’s important to understand the implications moving forward.
Plenty of things still remain unknown. For example, we don’t know the best approach to managing the seemingly opposing demands from the users for privacy and the advertisers for efficacy. However, industry players are putting ideas to the test to find a balanced solution.
What we do know is reporting and metrics will not be the same once Google officially does away with cookies in 2023. We also know that Google Chrome represents roughly 60% of browser use, while Safari, Explorer, Duck-Duck-Go, Bing, Firefox, and other smaller browsers represent the other 40%; and that 40% has already abandoned third party cookies.
As the dust settles in this space, a sort of philosophical discussion is emerging: what happens to attribution marketing, and are these changes going to send us back to the dark ages?
Prior to the advent of digital marketing, the only way we could get attribution was if the end-user provided that feedback by using a promotional coupon or filling out a survey. All the other sales and results were attributed to the general synergy of the media mix itself through regression analysis.
Digital marketing on the other hand gave us direct attribution: this conversion came from this lead which came from this ad served on this platform on this day at this time.
Neither were wrong for what tools we had available as an industry and what we knew at that time. But both are incomplete, and the privacy discussion is bringing that issue to the forefront.
The pre-digital way was missing accurate attribution, but with the one-to-many probabilistic approach, we did see the needle move. The digital way gave us accurate attribution, but with the one-to-one deterministic approach, we went too far over the privacy line.
The pre-digital way wasn’t tracking an impression so the assumption was that every piece of the media mix was responsible for the conversions. The digital way gave a lot more weight to the latter part of the cycle (either last touch or last non-direct point of contact).
The media mix model was all we had before digital came on the scene, but this is tracked over a long period of time to determine brand-specific trends in marketing efficacy.
The attribution model only works if you have visibility into who is seeing what and when regardless of which attribution model you use.
We are losing some of the visibility into attribution, but not all of it. But it’s also time to stop loading all the media spend into the bucket with the best attributable capabilities.
It’s time to modernize media mix marketing. Using the same principles and regression exercises to track long-term trends, we can likewise incorporate attribution metrics into the consideration set. Marry this with zero and first-party data, and you have a very healthy analysis for predicting future growth strategies as well as a better idea of who your primary target really is.
The fact remains: there is no silver bullet in marketing and advertising. In the coming months (and years) as government regulations continue to evolve, and service providers vye to ingratiate themselves to the general public as worthy of user trust, attribution visibility will wane.
By going back to the first principles of a balanced and synergistic media mix and integrating contemporary tools, like anonymous unified identifiers and zero and first-party data capabilities, we can get a more holistic understanding of the efficacy of our marketing and advertising efforts, while developing a conversational relationship with our end users.