Data privacy restrictions, the obliteration of cookies all thanks to Apple and Google have changed the measurement landscape and put the nail in the coffin for attribution.
Even with all the flaws of attribution the implementation was far easier and it was able to provide quicker results on what's working allowing for more flexibility. Compared to Econometrics and even Incrementality it requires time and a process to collect the right data
— DIPESH SHAH (@mrdipeshashah) March 20, 2023
Trusting the ad-platforms to evaluate their own homework was always asking for trouble. For those brands who had invested heavily into Facebook or Google attribution as their source of truth need to take back control.
For a lot of business trying to better understand performance and crack measurement without paying a significant sum they need to:
1) Understand their baseline
2) Understand trends
3) Understand what external variables are likely to have an impact and keep a track
— DIPESH SHAH (@mrdipeshashah) March 13, 2023
Google and Facebook announced they are moving away from attribution to focus more on econometrics. You can take it with a pinch of salt as both have vested interests, but it does clarify that it’s doomsday for digital platform attribution as they were the pioneers of pushing platform attribution with great success.
Why are Meta & Google moving away from digital attribution as a way of measuring marketing effectiveness?
Data issues are part of the story, but it’s bigger than that. This Google report gives some insight into their thinking. 1/9https://t.co/DObm8IJiMu#marketing#advertising
— Les Binet (@BinetLes) January 6, 2023
Econometrics and Incrementality has becomes the go to measurement solution for several reasons mainly it does not face the same data challenges that Attribution faced. They both complement each other and it’s about understanding what Econometrics and Incrementality can provide.
If you want to run incrementally successfully you still need a lot of the insights that come from Econometrics without it you are running blind
— DIPESH SHAH (@mrdipeshashah) April 20, 2022
Econometrics is a top down approach incorporating all forms of advertising that aggregates historical (min 2/3 years) data including external variables like weather, interest rates, promotion, and price etc to get a short to medium term view on how the marketing mix is impacting sales or any defined KPI. Econometrics delivers high level analysis allowing marketing teams to understand how the mix of channels and budget is driving sales but also the ability to look at potential scenarios helping to forecast and budgeting.
When marketing teams require more tactical and on-going insights this is where Econometrics struggles due to the lack of granularity of data and speed of delivery. Where the top down strategic insights are invaluable a lot of times marketing teams require campaign level insights. Econometrics works on the data it knows so when it comes to disruption to the brand, market, climate i.e., lockdowns it did not have comparable data.
This is where incrementality can be the perfect ally to Econometrics.
Incrementality looks at causation and the impact of exposure to a campaign against campaign KPI’s. Incrementality segments audiences into a test and control group and withholds media to the control group. The conversion rate between the two groups is the true incremental contribution to the channel, tactic, and campaign. This is one of the big advantages for incrementality the results will delivers insights where to eliminate media wastage and where to optimise budgets. The results are delivered in good time allowing marketing teams to be proactive and plan for the next test.
Incrementality proves its value by addressing the gaps within Econometrics.
Incrementality requires a lot of heavy data lifting, to truly understand the impact of the tests it needs a data scientist who can clean the data to generate the insights.
Both Econometrics and Incrementality complement each other, Econometrics is a top down strategic view to help with medium to long term planning, Incrementality is more tactical providing quicker results. Where it gets powerful is the incrementality tests can be used to train the Econometrics model helping with the long term planning and strategic view.
There is no right or wrong approach if you start with one only and then introduce the other at a later stage. There are many variables dependant on this: size of ad budget, revenue generated, skills set in-house or agency and most importantly time, resources and costs are a huge factor. If there is a choice between Econometrics and Incrementality then I would start with Econometrics as it provides a top down strategic view where the insights will be valuable across wider the business.