Volume 167: The Fallacy of Certainty.
The Fallacy of Certainty.
tl;dr Marketing is not, and will never be a hard science.
When Off Kilter started in 2019, we were firmly in the throes of a nonsensically ZIRP-fueled time. Brand purpose was touted as the “one true way” (it wasn’t). The goal of brand strategy was to “find your why” (nope again), and the idle musings of Gary Vee were heralded as profundity rather than the utter bollocks it mostly turned out to be.
As a result, the marketing science community's empirically researched and robustly evidence-based perspective came as an incredible relief. A moat of sanity surrounding a castle full of crazy, if you will.
In the years since I’ve often quoted How Brands Grow by Prof. Byron Sharp and additional marketing science-driven insights from the Ehrenberg Bass Institute and the LinkedIn B2B Institute. I’d even go so far as to say that the seemingly sudden emergence of marketing science-derived concepts into the mainstream of the marketing conversation (a mere fourteen years after HBG was first published) is one of the healthiest aspects of the current marketing discourse.
However, as marketing scientists and their disciples increasingly spread their gospel across social media and the trade press, it’s important that we take a critical look at the aspect of marketing science that I find most troubling - the presentation of a set of social science theories as if they’re hard science laws, which risks two inter-related problems - dogma and the fallacy of certainty.
In overly simplistic terms (hey, I’m not an academic), the difference between hard sciences and social sciences is twofold. First, hard sciences tend to represent closed systems where we can isolate variables and quantitatively measure them in a repeatable fashion (Newtonian physics, for example). Second, repeated experimentation tends to be very repeatably precise, leading to theories, quite rightly, being presented as laws. For example, we very precisely know what Earth’s gravity is; we know how much it changes from the poles to the equator, and we know this will be true every single time. Meanwhile, social sciences exist within complex, dynamic systems where it’s impossible to isolate every variable and where the underlying context often changes over time, which means repeated experimentation tends to provide fuzzy and somewhat differing results rather than precision. As a result, we shouldn’t use terms like laws to describe what is happening; instead, we should use terms like theories, rules of thumb, principles, and hypotheses.
This is important because when you present a rule of thumb as a law, it provides a flawed impression of the precision and certainty of the outcome and how repeatably it will occur under any and all circumstances.
Don’t get me wrong, I understand why marketing scientists present things the way they do. It’s a marketing technique in its own right. If the problem within marketing is an array of disparate and disconnected theories, anecdotal stories, survivor bias, and straight-up bullshit, then the way to cut through to a skeptical and quantitatively minded audience is to present your approach as empirical, repeatable, certain, and precise to the level of laws. In other words, wrapping marketing science in the cloak of the hard sciences is a carefully considered and deliberate exercise in brand positioning and differentiation. An observation marked particularly amusing by marketing scientists like Prof. Sharp, claim that positioning and differentiation are irrelevant. (My response is that if it’s so irrelevent, why are you so clearly doing such a good job of it?)
And there’s the rub. While the concepts marketing scientists present offer an excellent set of principles, rules of thumb, and hypotheses that we can use as a fuzzy starting point alongside an array of non-marketing science-derived marketing concepts, we must also be fearless in questioning what’s being presented as a hard fact when clearly it is not.
Take distinctive assets, which just turned Lyle’s Golden Syrup into a most unlikely topic of armchair discourse, largely among those who don’t appear to have much of a clue as to what they’re talking about but are, nevertheless, fervent converts to the church of Marketing Science.
Anyway, the idea that if a brand has unique, memorable elements to it, we’ll be more likely to remember them and thus connect the brand to a future purchase decision (if presented appropriately) makes perfect sense logically. And we have enough empirical evidence to support this supposition fuzzily enough compared to the alternative, which is non-memorable commodification, to place some considerable faith in it, in principle...
However, dive into the details, and we find ourselves in distinctly murkier waters. In research, marketing scientists will tell you that characters and mascots represent the most distinctive of distinctive assets. However, in practice, very few brands have these. So, is it the case that characters and mascots are always “most distinctive,” or is it simply that we notice and remember them because there are so few of them? Is it a law that a character will always outperform every other asset, or would this advantage decline rapidly if every brand piled in and followed the mascot playbook?
Equally, if you ask a marketing scientist when a rebrand is appropriate, their most likely response will be “never” because you’ll lose the meaning and memory imbued into your distinctive assets, resulting in nothing but a negative impact on your business. Yet, rebrands happen all the time, and I don’t see broad swathes of businesses suddenly declining and failing because of it. Yes, there is almost always a short-term drop in awareness and salience that occurs that sometimes has a negative impact on sales and profits, yet this tends to be a short-term rather than a long-term effect, especially for more sophisticated organizations that know what to expect and take steps to manage the change in order mitigate negative impacts. So, again, it begs the question of methodology. Is the marketing science position of “never” based on a true multi-year accounting of the impact of change across a broad swathe of rebrands, or is it based on short-term evidence, or is it largely evidence-free, and treated as an obvious extension of the distinctive assets theory. (eg. If distinctive assets are inherently valuable, then ergo, the effect of changing them must be negative). The truth is, we just don’t know. And, sadly, neither do most of the supplicants to marketing science who are so quick to pontificate on the subject.
I mentioned above that marketing science sees little or no value in either positioning or differentiation; however, as far as I can tell, this looks to be more the fruit of a flawed research methodology than the profound insight it’s presented as. Handily, however, it provides a great soundbite for differentiating marketing science from conventional marketing orthodoxy. Just take the Ehrenberg Bass Institute, which, as someone recently pointed out to me, “has expertly positioned itself as the only marketing institute on Earth doing ‘real marketing.’” Oops, they did it again. Damn, they’re good at this totally valueless positioning stuff!
So, while there is much to learn from and admire within the world of marketing science and much anecdotal nonsense that its debunked, what concerns me the most are its blindspots, and it seems to have a few. Largely, I think, because the underpinnings of marketing science appear deeply rooted in the dynamics of CPG/FMCG categories and how already large brands in such categories (with big budgets to spend on marketing science institutes, cough, cough) can leverage their brand strength to eke out marginal growth gains relative to competitors. As a result, if we’re not in the CPG/FMCG business and we are not already large, we must navigate the following blindspots:
It’s unclear that the “laws” of marketing science hold up 1:1 in non-CPG/FMCG categories. This feels most notable when we consider corporate brands that sit across multiple product types, services brands where so much of the experience is intangible, luxury brands where social signaling and emotional resonance matter so much, and technology brands where product innovation is so fast-paced.
It ignores qualitative information completely, relying solely on quantitative analysis, leading to some very real weirdness, such as discounting pretty much everything to do with the customer experience. This is one of the reasons I feel the dismissal of positioning and differentiation is more about a methodological blindspot than fact.
It does little or nothing to reflect how organizations utilize brand strategy, positioning, etc, to guide internal decisions around marketing communications, product innovation, customer experience, etc. The glib statement that “your positioning is whatever your last ad says it is” is useless if the net result is chaos.
It fails to differentiate between how a very small company with extremely limited resources might grow its business relative to how a large company with vast resources can incrementally leverage its brand for growth.
It largely dismisses the bottom-of-funnel tactics we commonly label “performance marketing” completely. While I agree that these tactics tend to be much less valuable than typically presented, it would be equally incorrect to view them as being valueless under all circumstances.
Now, this is simply my opinion. I’m a magpie for ideas and principles and love rules of thumb as starting points. I have no problem at all living with conceptual ambiguity, see little or no reason to quantify everything, and am quite happy to admit that non-quantifiable qualitative data is often the most valuable. I also think plenty of marketing theories that precede marketing science remain important.
So, I apologize to any marketing scientists out there who might think I’ve mischaracterized your field through my ignorance. I’m fundamentally not anti-marketing science; however, I am fervently anti-dogma and deeply cautious about the dangers of false promises of certainty.
What matters is to think critically about what you’re doing, run your own experiments, gather your own evidence, and use the thinking of others as principles, rules of thumb, and start points. But don’t be afraid to find your own way after that, make up your own mind, and discover your own evidence for what works in your situation, category, and company. And, especially, to track how that changes over time as the impact of competition forces diminishing returns to kick in.
Because that’s what really matters at the end of the day.