“Modelling” seems like the new “Digital” in today’s marketing landscape; a term bandied around with little understanding. Similarly to “digital”, it’ll likely become a differentiator for brands and agencies, before it becomes ubiquitous. This article however, dives into the scope of the problem, and the resultant opportunity for brands.

How many organisations refer to buying media as investments, outside of the marketing department? My belief is very few, as it’s commonly seen as a cost of doing business, rather than an investment for growth.

When asked about the role of modelling in a post-Covid world, Nina Bibby, CMO at UK telco O2, explains in the article how econometric modelling is the only way for them to truly understand on a large scale, the drivers of growth, what contributes to a sale, and how it helps accelerate their media investment profitability.

The O2 use case is a key distinction that (usually) separates brands who use modelling from those that don’t.

Modelling, econometrics and otherwise, has been used for many decades, and often shrouded in secrecy and complexities in a black box. Only in relatively recent times is it being increasingly  adopted by the marketing world.

There is now enough data from businesses, campaigns, and from the wider competitive environment, to use data as a highly effective strategic asset through modelling.

And as econometrics was originally applied to mostly the finance industry, informing investment decisions, the trend is emerging for sophisticated marketers who “wargame” campaign scenarios to predict things like sales and brand uplift with a good degree of certainty, managing an “investment” rather than simply a “budget”.

This trend is likely to become a necessity as businesses enter a post-Covid world much leaner, with marketing budgets to match.

Trends like zero-based budgeting will only exacerbate the crunch, meaning that every dollar of marketing and media spend will need to be proving its effectiveness over both the short and long term.

Modelling tends to drift towards the short term at the moment, looking at campaign level issues like multi-channel attribution for sales.

The data is substantially easier to gather and model, and doesn’t rely on probability. While this is certainly useful, the example from Adidas is timely - breaking every rule in Binet & Fields The Long and the Short of It, focusing too much on these short term considerations can skew how you evaluate your media, and leads to traps like collinearity.

Channels like SEM are rated much higher in their contribution to a sale than they should be; the purchase intent from the consumer is already there, but they’ve happened to click on your ad while trying to buy the product.

Brands who are serious about modelling, and want to get the most out of it, need to be looking wider, and longer term. Blending media data with econometric data is the logical next step.

Media mix modelling gets part of the way there, but tends to be static, expensive and takes a long time, though advances in cloud processing of data are making this easier, faster, and much more affordable.

And when you get it right, you’ll have the fuel to not only interrogate creative briefs and media plans - but P&Ls, with the insights to land you a seat at the table, defending and growing budgets, and proving your marketing investment’s contribution to your company’s bottom line.