Multi-channel media attribution and other models are quickly replacing traditional cross channel marketing models. Which is the best one?
In my last post, I wrote about marketers investing more in cross-channel marketing and how they should be wary to make sure their investment in a custom multi-channel attribution model is equally important.
I thought I’d follow that article up talking about the types of models that are available, and how marketing offices should go about choosing the right one.
Cross-channel marketing modelling has traditionally been synonymous with “top-down” marketing mix models.
In recent years we have seen the appearance of “bottom-up” attribution models due to an increase in software and technology available to marketing offices.
Each campaign must be developed and implemented on the merits of its own business and marketing objectives, and no cross-channel marketing model is perfect for each campaign.
I have long argued that customisation of marketing and attribution models is far better than dealing with a static model.
Kohki Yamaguchi over at MarketingLand has a very good writeup on some common cross-channel marketing models we come across.
He highlights three of the most common approaches to model cross-channel marketing campaigns:
Econometric top-down approach (e.g., marketing mix models)
Algorithmic bottom-up approach (e.g., attribution models)
Machine learning approach (e.g., agent-based models)
A key differentiator across the models will be real-time simulation and prescriptive capabilities; how well can the model provide firm granular data as well as offer predictive analysis.
What is common to all of these models is data. It requires the right technology and the right business processes to maximise the technology. As Ashley Smith states in an article for searchCRM:
“The software is important, but it can’t make up for bad data, lack of clear marketing objectives or poor workflows.”
It requires working with many departments across an organisation in developing common business and marketing goals, and implementing practices that will ensure all facets of the organisation embrace the marketing model.
Effective cross-channel marketing modelling means returning to that age old argument of breaking down silos within marketing offices and the wider organisation.
We’ve been hearing this for awhile. So why hasn’t it happened yet?
