Applied Thinking: Flaws - James Devon planning director, MBA

We strive for perfection. We want things to be lined up nice and tidy; understandable. So we measure, analyse and evaluate. We talk about virtuous cycles, failing fast and learning faster. Correctly, we're looking for evidence on which to base decisions, to prove to the stakeholders that our plan is working.

Applied Thinking: Flaws - James Devon planning director, MBA

There are flies in the ointment, though; flaws blemishing our quest for perfection. We're human, our customers are human, and our metrics suffer Schrodinger-esque tyrannies of measurement. Pristine never is; clarity has shades of opacity. The flaws are often minute scratches that give us the confidence to proclaim the evidence as untarnished. But the best super-optimised-universal-analytics-trigger-based-media-mix-econometric- attribution plan doesn't work perfectly.

Rather than worrying about flaws, endlessly tweaking and drilling down, embrace them. Do your best to mitigate them, but also work out how to deal with them, find proxies and, arguably most importantly, make sure you don't lose sight of the meaning of the whole.

In a twist on Aristotle, the psychologist Kurt Koffka stated: "The whole is other than the sum of the parts." You know the black-and-white image of a Dalmation dog in dappled light? You perceive the whole of the dog, but without the parts that normally make up your impression of the dog. This is a useful concept for working around marketing's flaws.

The increasing number of metrics, combined with the pressure to be accountable, leads to too much focus on "parts" metrics - often short-term diagnostics that aren't necessarily indicative of long-term success. Optimising the parts is essential. However, when we're trying to shift business metrics, will a 0.01 per cent improvement on a display ad make a fundamental difference when the programme in its entirety isn't delivering ROI? Look at the elements of the system, but look at the system as a whole too. Numerous tales are told in which seemingly underperforming elements are removed to the detriment of the whole.

In MBA's work with the cross-industry social measurement initiative #IPASocialWorks, we frequently find brands focusing too much on what they can count - the readily available numbers (the parts) - without connecting the effects on the wider business or marketing objectives (the whole). For social to really be taken seriously, we need to know not only that it works, but also that it is more effective than the other available options.

Another flaw is the assumed linearity of purchase decisions. It's much more complex than that as options are added, mulled over, rejected and chosen in a seemingly chaotic fashion. We envisage the customer decision as a dynamic, digitally fuelled ecosystem to account for this nonlinearity - a concept that helped Sage by Heston Blumenthal achieve dramatic results in its launch year.

There is an inherent flaw in outbound communications that assumes the defunct linearity. Most people aren't in the mood for listening to you even subconsciously (although, hopefully, some positive emotional trace remains), no matter how insightful the proposition and well-optimised the targeting. We must make allowances for the "inbound" so that people can find their way to your brand and the content they need to choose you.

Academic research looks for conditions that are necessary and, together, sufficient for something to occur. Applying that to the "inbound" requires optimising the search and content strategy together with the user experience of your website. Technical developments such as universal analytics mean that more "whole" evaluation can be applied in addition to optimising the parts (eg. with Everest, we'll be linking high-value customers to specific search terms). "Sufficient" is a particularly interesting notion as it doesn't necessarily need to be exhaustive information. Seeing the right parts can create a sufficient perception of a whole - draw three Pac-Man eating the corners of an invisible equilateral triangle and you will perceive the triangle.

Flaws exist - we're dealing with the complexities and vagaries of human psychology. They're annoying, but needn't be paralysing. They should be acknowledged and ignored. Flaws are there to be worked with and around, keeping in mind that the flaw is only a part of a whole. More often than not, the whole is the objective. And it is other to the flawed parts.


The thought that should keep marketing directors up at night is ... How do I connect data together so I have a chance of knowing what's really going on and how this relates to my business KPIs?

The thought that actually does is ... As above - this is a real and present issue.

Great thinkers who have inspired me include ... Professor Ian McLaren (cognitive psychologist), Dave Birch (on the future of money) and, rather ubiquitously, Steve Jobs.

The idea that excites me the most (professionally) is ... There are many, but stealing Dave Brailsford's concept of "the aggregation of marginal gains" is a good one.