The phrase “economical with the truth” was famously uttered by Sir Robert Armstrong, the then-Cabinet Secretary, in 1986 during the Spycatcher trial. It means something falling short of a lie, but a fair way away from truth. It’s a good phrase to keep in mind next time anybody tells you that TV returns £1.80 ROI. The truth is a whole lot more complex than that and it’s usually shopper marketing, or more widely what used to be called “below the line” that loses out.
Some problems with econometrics
Econometrics is at best problematic for FMCG brands. Like many aspects of the marketing process it was designed for brands that own the sales process and relationship with their customers so takes no account of the fact that with FMCG brands there is this entity called ‘the retailer’ that controls distribution, price, promotions etc. The reality is that any one of these factors on their own will influence the sales of a brand not just by a single-digit percentage but by a multiple factor.
Given what we know about the mutliplicity of factors that drive sales performance in FMCG, it’s striking how often, when reviewing econometrics work, the data set used was not split by store, did not include secondary display information and did not have an availbility overlay. It’s difficult to see how any resultant model can be truly predictive if it doesn’t at least attempt to replicate the reality in stores, hugely complex though this may be.
Beware simple conclusions
Even assuming that the study has been prepared with the correct level of input detail, the output often seems simplified to destruction. “We know that when we do TV it returns £1.80 ROI” (for some reason it’s always £1.80) I’ve been told by several clients. Really? Think of all the variables involved in TV - not least how keenly space is bought - and it’s obvious that whatever the figure is, it cannot be the same across all possible uses of the medium.
Part of the problem is senior management craving simplicity and clarity. I’m all in favour of simplicity and clarity but sometimes the world doesn’t work that way, and it’s obvious that marketeers have a vested interest in “averaging up” when it comes to the performance of their department or indeed of their own brands and agencies. And the simple (positive) headline is the one that sticks.
A new approach
So where do we go from here? Brands should apply the maximum amount of scientific method to understanding the effects of their marketing that’s for sure. But beware projects that are so complex that no one can explain them. How objections or clarifications are pitched is crucial - don’t say “does anyone actually believe this?”. Do say “Do we feel that we really understand this methodology in detail?” (the answer is almost always ‘no’). Follow up with “I’m wondering how much weight we can put on this given we’re not sure how the conclusions are calculated?”.
However there is plenty we can do. Shopper is the most measurable marketing activity on the entire plan, given that it takes place at purchase points, and given that almost all FMCG is impulsive in nature. Measurement of activation rates is trivially easy using store-level EPOS data and control groups can be generated for almost any activity.
We can take a leaf out of digital, which has rapidly become the dominant marketing channel without being able to demonstrate, as far as I can see, any ROI at all for FMCGs. I am yet to see a single study that demonstrates that the clicks we’ve paid for caused the purchases, rather than simply occurring just before them.
It’s easy to forget that the universal dimension of all marketing activity is reach. Shopper is great for reach as media becomes ever more fragmented. In the sea of hype about personalisation it’s also very easy to forget that the essence of FMCG is mass-market - if ‘personalisation’ were re-branded as ‘audience fragmentation’ it would probably help us to think more clearly.
Shopper is one step ahead of most other channels in that we actually do understand short term ROI very well. The next step is to connect our understanding of reach with activation success. Then we can approach the holy grail of true ROI understanding and plan shopper against audiences like TV is planned but with the closed loop of total visibility of results. In the meantime, if anyone tells you they understand the precise ROI of all marketing spending or all TV spending - be sceptical.