Paul Baker is a Marketing Mix Modeling Guru and a Respected World Expert in Media & Marketing Effectiveness. Paul is the former CEO of OHAL for over thirty years, he had been appointed in 2001 as Global CEO of OHAL.

Coping with COVID-19

How is the recent COVID-19 outbreak impacting your day to day work life and what steps are you taking to cope with this?

The recent outbreak didn’t really impact my day to day work life, but it created an opportunity to provide my close network with critiques on the various research papers currently being produced.

Among those, I find that many interesting papers can be found at Oxford University and Imperial College, and while they did a great job, I believe it would have been very useful to also include a break-down of the insights by urban areas and by actual timeline.

My advice is that if you are a modeling agency and you are currently facing a downturn and want to bring value to your team, you can start an internal initiative to encourage the team to analyze the major papers that have been produced by the above-mentioned institutions regarding COVID-19.

Impact of the Pandemic on Marketing Measurement

According to you, how will the COVID-19 outbreak impact Marketing, Marketing Measurement, and Marketing Mix Modelling in particular?

Looking at the ecosystem now, I would say that each industry has been affected in a different way.  In most industries, there is little point in including the virus time periods in the data set to be examined.  Probably the worst thing is to include it and then use some form of surrogate ‘dummy variable’.  A major skill will become in defining when a more normal period begins.

  • For some companies, it can be possible to identify the base to end 2019, and then see how a new base settles once normality is restored. This needs to be handled carefully (statistically).
  • At this point, another skill that should be employed is the ability to develop hypotheses to be tested. There will be a great temptation to throw the data at established models and see what happens.
  • Many in the modeling industry will be telling their clients that it is the time to invest and so should the modeling companies. And for this, it is critical to developing better modeling techniques, and by this, I do not necessarily refer to statistical tests, or software packages but refer to the ability to answer the clients’ questions based upon more reliable and objective measures.  For example, find something better than ad stock or realize that R-squared is not a particularly valuable measure.

Marketing Measurement and the Economic Recovery

How do you think Marketing Measurement can contribute to overcome the economic consequences of this crisis?

The current dilemma that we might be seeing is that many modelling companies want to show that there is a great ROI from marketing expenditure. Most clients are commissioning from within the marketing department and want to justify what they are doing while modelling agencies want to keep clients happy and get good ROIs

Despite this caveat, models can definitely help clients and it should be possible to provide answers that improve the decision process at this critical time, timescale is particularly important as companies are going to need to make major decisions quickly.

Many of these decisions are well above basic marketing requirements and will affect the whole core of the business.  For market models, it becomes essential that answers are ‘Right rather than Wrong’ and that perfection would be considered as the enemy.

An absolute example is the modelling currently going on for Coronavirus, if given as a project, most companies would say that 2-3 months should be reasonable. However, groups like Imperial College had 2-3 days to produce something more Right than Wrong, and so far, it would appear that they have largely succeeded.

I believe that there will be few cross-industry conferences and probably some cross-industry publications. However, due to the competitive nature of the relationship between industries, it might be very challenging to collaborate.

How the Marketing Mix Modeling Community can Cope with the Pandemic

Any advice you would like to share with the community to cope with the COVID-19 outbreak?

I advise everyone to listen to governmental advice on the best practices to slow the COVID-19 outbreak and stay healthy.

For companies or agencies who are facing downtime and have some of their team members underemployed, I would encourage them to focus on improving their wider business and marketing skills. This might be the best time to improve their research skills using both quantitative and qualitative methodologies and to strengthen essential business and statistical skills (e.g. automatically calculate sampling error in tracking Studies, design statistically valid tests.)

At the end, I would say that many academic papers may lack depth, partially due to the absence of corporate data that usually remain confidential. However, it is still important to stay up to date with decent trade magazines publication to enable us to extend our knowledge.