Econometrics is a statistics-based methodology to quantify the relationship between cause and effect: in other words, a ‘scientific’ way of identifying the impact of different marketing and media activities on consumer behaviour. The aim is to create a mathematical ‘model’ which, in the case of marketing, is able to attribute the contribution of each activity to sales or any other tracked outcome.
Other phrases that are used or associated (almost interchangeably) with Econometrics are ‘Marketing Mix Modelling’ and ‘Time-series regression’.
Some agencies/consultancies will, by giving it another name, claim to have invented it – but econometrics is a statistical discipline taught and studied in universities and colleges, and not invented by agencies.
Why you should use/carry out/be aware of Econometrics
The technical diagnostics that go with econometrics require a trained statistician to properly interpret but a more common sense interpretation of how the model relates to the real world is just as important. Econometrics is a technical exercise but it should not be just a technical exercise – That way lies madness and spurious correlation in equal measure!
All of which is a shame, because econometrics, properly applied, is a really useful tool to help marketers understand what has worked and what hasn’t.
In a world where the lure of the shiny and new is ever more seductive to the enthusiasticmarketer, and where marketing accountability is a business imperative, knowing where to invest time and money, rather than relying on the latest spin, is of paramount value. After all, the media agency business model is fundamentally reliant on clients spending more money with them on the media mechanics which deliver the most profit: hard as it may be to believe, these are not necessarily those which are in the best interest of the client.
Who uses Econometrics?
Econometric modelling is generally commissioned by one of the following: Marketing/Media, Research or a specific Marketing Effectiveness function.
It is used – and indeed useful – as a means of determining where to concentrate marketing and media investments.
It is increasingly referenced by the Finance function within an organisation as a business case proxy for marketing investment.
How to get the best out of Econometrics
Quite simply, econometric modelling is poorly used if it is merely a post-campaign ROI justification tool.
Far better is the use of econometric modelling to identify diminishing returns and provide the basis for future-facing planning, budget optimisation and forecasting at the STRATEGIC PLANNING and the CAMPAIGN PLANNING stages. Econometrics should be part of a ‘Do-Learn-Do’ culture of continuous improvement.
Of course, the models therefore need to ‘make sense’. For this, you need an experienced modeller, who combines the necessary ‘stats’ knowledge with sound business/marketing/commercial – and common – sense.
Automated modelling systems do exist and they are attractively priced, with apparently similar outputs. Our experience is that they require significant manual intervention to make the outputs truly accurate, as opposed to apparently accurate.
Costs and Benefits of Econometrics
Projects are typically costed on a ‘per model’ business, based on the amount of time taken by an analyst/consultant.
The benefits are simple:
- Empirically informed decision making (not based on ‘wooly’ metrics)
- Optimised budgets
- Identification of monies which can be diverted to channels with better returns or which can be invested in new initiatives
- Better forecasting
- Higher ROI over time