The Real Issue is Transparency, Not Technology
There’s been a fair bit of noise recently in our industry around how the advent of automation and/or AI has potentially diluted the quality of our work via access to “cheap”, fast modelling processes.
Whilst this may undoubtedly be true in some cases, automation isn’t the villain of the piece: transparency, or rather a lack of it, is.
There’s been a fair bit of noise recently about how the advent of automation/AI within the effectiveness industry has led to an apparent dilution in the quality of the underlying measurement models that are used to measure the effectiveness of media.
Whilst I’m sure no-one would disagree that using a bad model to inform media strategy would likely result in a bad outcome, the source of this problem is not down to the level of automation used but rather down to a lack of transparency in our industry in general around the quality of the underlying modelling approaches used.
Very often practitioners in our industry cite the supposed “IP” of their approaches as a reason for not being able to open up their models to independent 3rd party scrutiny. However, the fast paced nature of the marketing industry and the need to simply produce results, when coupled with an almost dogmatic view that marketing mix modelling could, and should always be done as part of a largely manual, bespoke process creates a culture of corner cutting that ultimately leads to bad modelling approaches being used.
Key to Marketing Efficiency and Quality
Automation, therefore, isn’t the problem but rather the solution as it enables high quality modelling and analysis to be carried out speed and indeed at scale. The adoption of automation does however need to go hand in hand with greater transparency, with practitioners opening up their work to independent 3rd party scrutiny as a means to guarantee the quality of their work.
Indeed, automation can bring other benefits as well, not least in a reduction in the time taken to carry out the work and therefore ultimately a reduction in the cost to the client, but also by effectively allowing for more focus to be placed on the insights that are generated from the modelling process rather than on the creation of the models themselves.
As effectiveness practitioners, clients ultimately look to us as experts who can help them make their marketing more effective. Therefore perhaps we should place more of an emphasis on how we can better help them to achieve that goal, rather than on disparaging the innovation that our industry so desperately needs.
