These are interesting times for econometricians.

The models we build are based on relationships. Relationships, chiefly, between sales and the things that drive sales – advertising, promotions, seasonality, distribution, brand equity, trends in consumer tastes, habits and preferences and so on.

In fact, we couldn’t do what we do without presupposing the existence of these relationships. What would be the point in measuring the effect of TV on sales last week if, next week, that effect is completely different?

So, as people are starting to ask how Covid-19 has changed the world? This amounts to asking how have those relationships on which we built our existences, or at least our marketing plans, have changed? There are only 3 answers to that question: (1) loads, (2) somewhat or (3) not at all.

How have those relationships on which we build our marketing plans changed because of Covid-19?

Possible answer 1. Loads

This means that the relationships that formerly existed between such things as TV and sales, or paid search and orders, or seasonality and enquires have fundamentally altered and we can no longer base our plans on them.

Possible answer 2. Somewhat

If this is the case, it means there have been some changes at the margins – an ROI up here, a t-stat down there – but fundamentally the world is much the same as it was B.C.

Possible answer 3. Not at all… well, you can guess what this means.

So, how do we find out what the answer is for our company? The exciting thing for our breed of geeky marketers who are obsessed with accurate ROI figures, is that we’re moving into the time where we have actual data to analyse and model to help us determine how relationships have changed due to Coronavirus, and how we need to adjust our thinking and maybe even worldview.

Let me emphasize that again: actual data i.e. NOT speculative guesswork based on our particular, peculiar parcel of prejudices that would normally be pontificated over whilst procuring a pint in the pub.

With that in mind, here are our best tips for how to move your marketing forward with some timely measurement.

Moving your marketing on post-Covid

  1. Do Some Modelling

Yes, yes, we would say that, but there really is no substitute for hard facts when it comes to making decisions. You can speculate as to how Covid has, will, or won’t affect your business. Much better to find out, and you’ll only know by updating your econometric models. And for the nay-sayers in your organisation: yes, you can model it. We can measure the things we know and get a read on them. This helps us understand the stuff we don’t know.

  1. Do it at the right time

Whilst we empathise with clients’ fervent desire to know stuff NOW, there is no point in doing any sort of modelling until you have enough data to evaluate how Covid has impacted your sales.

The right frequency will vary by category. For some clients, we’re already building models that quantify the Covid effect. In other sectors, with complex purchase cycles, you may have to wait a while. Also, don’t just be led by the desire to know – but the need to make decisions. If you’re not planning on doing any advertising for the rest of the year, then why not wait a little longer for more data to come in before you do any modelling?

  1. But before you do any modelling, look at your data

Just eye-balling your data alone should give you a feel for what effect, if any, Covid has had on your sales. If the chart of your sales sails gloriously on, apparently unbuffeted by the vicious winds assailing it, then there’s every chance that your advertising effectiveness will also be unchanged. Of course, it won’t give you the full picture, and you may have to, as we did, deal with offsetting negative and positive factors (i.e. Covid and a TV campaign), but it’s a good starting point and where we always start when we’re asked to model a business.

  1. Bear in mind the difference between marketing and non-marketing drivers

Statement of the obvious: Covid’s effect will differ markedly between different brands and sectors. HOW it impacts will also vary. It is perfectly possible for Covid to have significantly dented base (i.e. non-media/marketing) sales whilst having little or no effect on the effectiveness of your advertising. We’ve witnessed this already, modelling a business that, once the negative Covid base effect was controlled for, had a very successful TV campaign. An effect on sales is not the same as saying that advertising effectiveness has fallen.

  1. Keep an open mind

Covid has hit sales. Has it? Or was it the fact you pulled your advertising at the outset? How did natural seasonality affect your results? Even if it has dented sales, what is the profile of the effect? Was it big to begin with as everyone ran for the hills, SUVs laden with bog rolls, and then tapered off, sales recovering until they were practically normal? Perhaps, after the initial shock, Covid actually boosted sales, as people sat at home, surrounded by mountains of bog roll, with nothing better to do than buy stuff on the internet?


As we move on and forward, understanding your data and what it shows you is as crucial as ever. If you’d like help or advice on how to do this, then let us know. We’re happy to have an initial chat, give you some pointers, and if we can help take it further.

Philip Gaudoin

Sign Up