When is evidence of marketing uplift not evidence?

These days we can access data at any time: clicks, website visits, likes, shares, search, live sales by minute, by second… but how do we make sense of this data? Why are uplifts in these metrics not the same as campaign success? As a marketer you must have heard the old adage that correlation does not mean causation. But what are the risks and implications of making big decisions based on these types of data and simple correlation? We need to make sure that there are in fact evidence of marketing uplift to justify spend.

When data moves in the same way it does not necessarily mean they are related to each other. Spurious correlations can be detrimental if not spotted – or if you like, is indeed spotted, but not recognised as such. Specifically the case where two metrics move in the same direction, but not because they are interacting, but rather because they are both caused by a third underlying factor or just happen to occur at the same time. For example, as ice creams sales increase the rate of drowning also increases which therefore must mean that the consumption of ice cream causes drowning – in fact, the true driver (the latent factor) is summer seasonality.

So how can we make sure that these faulty correlations are recognised for what they are?  Well, when assessing the success of activity there are a few factors we must consider like for example seasonality. What types of activities fuel other activities such as promotions, world events, social media, mail order and other forms of advertising? Are there activities promoting new product line launches or perhaps expanding on an existing product like creating a new flavour perhaps?

What are the results of misinterpreting data? Misallocating resources is a real possibility by using media platforms that simply do not work. Back to our swimming example: imagine trying to stop drownings by outlawing the sales of ice cream!

This is why econometrics is so powerful and at BrightBlue we take on a holistic approach to modeling because we know the pitfalls.

 

Brightblue Consulting are a London based consultancy which help businesses drive incremental profit from their data. We provide predictive analytics that enable clients to make informed decisions based on data and industry knowledge. Through Market Mix Modelling, a strand of Econometrics, Brightblue has a proven track record showing a 30% improvement in marketing Return on Investment for clients’ spend. If you are interested to find out more please contact us through email by clicking here and one of our consultants will get back to you shortly.