According to Investopedia.com, Econometrics is, “the use of statistical and mathematical models to develop theories or test existing hypotheses in economics and to forecast future trends from historical data. It subjects real-world data to statistical trials and then compares the results against the theory being tested.”
So, there you go.
We understand not everyone has the same level of knowledge of Econometrics as we do, so we’ve written a two-part, 10 step guide to understand:
- What Econometric Modelling is
- What it isn’t
- How to use it
- When to use it
1. What’s it called?
Econometrics can come in many different guises. You may have heard it referred to in the following ways:
- Econometrics
- Mix Modelling
- Marketing Mix Modelling
- Media Mix Modelling
- Econometric Modelling
- Marketing Attribution
- Multivariate regression modelling (a personal favourite)
In essence, these tend to refer to the same thing. For simplicity, we favour “Econometric Modelling” or just plain “Econometrics.”
We try to avoid calling it “attribution.” Whilst econometrics does indeed attribute sales to sales drivers, that word is generally used to mean something different in the context of digital measurement and, frankly, just confuses everybody. It would be easier to say that econometrics is one means of marketing attribution.
Confusing digital attribution with econometrics is a bit like doing this
2. What’s it for?
As marketing teams juggle mounting demands with limited resources, the age-old question of where to channel those precious resources arises. To make sense of the data, Econometrics is the go-to tool for unravelling the influence of external and internal factors on a specific KPI.
However, it’s important to remember that Econometrics is just one tool in the decision-making process, to be used in conjunction with other research and business knowledge. Insight professionals are skilled at synthesising this information into actionable recommendations, but econometrics alone cannot provide all the answers. What it ISN’T is an all-knowing computer model that will tell you what to do with accuracy and certainty. Sorry, but it needed saying.
It’s crucial to strike a balance between over-reliance on Econometric Models and complete dismissal of their value. While some situations may not be suitable for modelling, in many cases, Econometrics can provide valuable insights that complement intuition and gutfeel. By avoiding the extremes of blind faith or dismissal, marketers can make informed decisions and allocate resources effectively.
Econometrics can answer many questions, and we’ve outlined 5 key uses of Econometrics in our blog ‘5 uses for an Econometric Model’.
Econometrics: Not a silver bullet
3. How does it work?
Econometrics as we use it is a statistical process that involves analysing sales data to determine the effects of various factors such as media, pricing, promotions, and more. Market Mix Modelling (or insert other term mentioned above), encompasses all these factors. Even if you’re only focusing on media, you’ve got to keep tabs on the other factors that could be driving sales. Otherwise, you might end up with some wonky results.
Econometrics uses a fancy technique called multiple regression to track how things like ads, prices, and even the season affects sales. And it doesn’t stop there! We also run a series of tests to see if these factors have a real impact on sales with a confidence level of 95% or higher, something you may have heard referred to as “statistical signficance.”
Some maths – pure, not statistics, but who’s going to know?
The model doesn’t play favourites with data. It doesn’t care what factors are being analysed. Whether a particular initiative was a pet project of someone, whether an ad went viral or even won an IPA award, the model remains objective in its assessment. Its sole focus is on determining what happened and what consequences followed, without any personal biases or sentiments.
4. What does it do for me?
Econometric Modelling is all about measuring the real impact of your actions at a high level. The trick is to focus on incremental impact, which means only measuring the extra sales you’ve gained. For instance, if you run a price promotion, you’ll naturally sell more, but not all those sales will be extra. And in the world of paid search, if someone was already going to buy something, sees an ad and clicks on it, that’s not incremental. So, to get the real picture of what’s working and what’s not, Econometric Modelling measures incremental impact, that is the extra sales that you got from doing X or Y activity.
Price promotions: We’ve all taken advantage of a price promotion. The question is – what would we have done otherwise?
What do we mean by “at a high level”? Econometrics is a superhero when it comes to measuring the big picture. But, if you’re itching to know which of your 75 digital banner ads hit the jackpot, Econometrics is too zoomed out to give you that level of detail. Don’t fret, though. There are other tools, like digital attribution modelling, click-through analysis that you can use in conjunction with Econometrics to help you answer those granular questions.
5. What it does well
Econometric Modelling is great at giving you a single view of the effect of all the different potential drivers in one analysis. Often, we have pricing pros, media gurus and digital wizards each doing their own analyses in their own way, leading to, at best, highly skewed results. And who’s even looking at the weather? Probably no-one.
Enter econometrics! It levels the playing field by comparing all aspects of business equally. That’s often why the Managing Director is often the most eager to get their hands on the results.
That’s our Part 1 on the essentials of econometrics. See Part 2 for what econometrics does badly and what you need data wise.
In the meantime, if you’d like to dig a bit deeper into Econometrics, download our free ebook ‘Guide to Econometric Modelling for Modern Marketers’.
Tom Lloyd