Digitalization has created a new world of advertising: The journey from a customer’s first contact with the brand to the purchase can be tracked in detail. Marketing attribution allows us to distribute budgets to the most important online channels and publishers for better online marketing.
On the other hand, however, marketers still use classic marketing mix modeling, which allows them to see offline marketing channels from the top and determine which channels drive most sales. Budgets can be adjusted accordingly. Marketing mix modeling can also be implemented to simulate the effects of budgets being shifted from one channel to another.
MARKETING MIX MODELING VS. DIGITAL ATTRIBUTION: MACR0 VS. MICRO
Both, marketing mix modeling and digital attribution, have their advantages and disadvantages. Roughly summarized, they do the following:
Marketing mix modeling measures the sales impact of different marketing channels retrospectively, for three months up to one year. It considers offline channels as well as factors independent of marketing such as economy, trends and seasonal influences.
Digital attribution measures the contribution of each touchpoint on a customer’s purchase decision across different devices. That way, we can view the customer journey in such great detail that we can focus more on individual touchpoints in online marketing.
While the approach of marketing mix modeling provides historical data to view marketing investments, digital attribution goes into greater detail so we can see our marketing activities from the perspective of individual customers.
ONE MODEL ALONE IS NOT ENOUGH
The different perspectives can have one negative effect: With both approaches we might overlook some of the influencing factors on our sales. While marketing mix modeling allows us to shift budgets to the most promising channels, digital channels only have limited insights. Neither do we get information about how exactly the customer journey took place via the different touchpoints, nor do we know how other online measures affected the purchase decision. We need the user data from the digital attribution to answer these questions.
On the other hand, we cannot tell from attribution if there were any external factors that led to the first touchpoint between the customer and our brand. Baseline sales are not included in digital attribution. It seems as if each (positive) contact with our brand could be rated as a success for our online marketing activities. But there are numerous products, that owe their success to external factors such as buying power, seasonality, trends and so on. Marketing mix modeling is the only method to find out how (and if) advertising affects sales positively, because it distinguishes between sales that would have happened anyway, and those we promoted through push marketing efforts.
Below is an overview of what marketing mix modeling and multi-touch attribution can and cannot achieve.
MARKETING MIX MODELING AND ATTRIBUTION SHOULD NOT BE MIXED RANDOMLY
In order to bundle the advantages and balance the disadvantages, many companies roughly combine both models and work with two different sets of data – one set of macro data from marketing mix modeling and one set of micro data from digital attribution. But that might lead to misinterpretations rather than complementation, which could end up causing more damage than good. A makeshift combination of both models is no solution either, especially because it means double effort.
HOLISTIC INSIGHTS AND OVERVIEW THROUGH MARKETING MIX ATTRIBUTION
The solution is a combination of both models in one single set of data, as in marketing mix attribution. This approach creates holistic insights and an overview of the different channels and their correlations and impact on consumer behavior. Offline campaigns are no longer recorded separately, but put into context with the user data that was collected with the attribution approach.