It seems a week can’t go by without people losing their marbles over some new thing: abstract, techy, quasi-predictive, and always, always hyperbolic. Well, here we go again. The LinkedIn bros are hyperventilating about how our personal bots will be shopping for our new car, speaker-system, gumboots, and tampons.
“The old commerce world is dead!”, they scream madly into the night, wide eyed and trembling in excitement, or angst, or both. Their doctor scrambles for a benzodiazepine. They ignore her. Brand marketing is over! You can throw away your degree. Distinctive assets? Your bot won’t care. And customer management? Gone like the wind, mofo.
The die is set, they reckon.
Robots will buy from robots, chuckling maniacally as they haggle over price, akin to a drunken scrap at the Hong Kong markets. All for you, of course. Then they’ll crack open a cold one, scratch whatever bots scratch, and finish booking your rectal exam. Yet, by their grace, we shall remain. They’ll keep us in a huge strawberry jam jar beside the virtual fireplace.
Versions of the narrative are everywhere. The recently released, ‘Martech for 2026’ report claims that agentic shopper bots represent “disruptive change”.
But does it?
Let’s explore
I’ve been around. I know you have too. We’ve all seen similar declarations: The death of TV, of newspapers, and of taxis, and the arrival of the metaverse as the trans-disciplinary revolution of planet earth where our avatars abscond with our souls and flirt on our behalf. Because, you know, who wants to deal with rejection? Don’t worry, they assured you, it won’t hit on your Mum.
But whether the narrative is breathless, or slightly more measured in tone, the thing to understand about most commentators on the subject of, ‘agentic’ is that they tend to exhibit an almost total ignorance of actual marketing, and of managing a customer base. In all the excitement, they risk proclaiming a disruption or a revolution to a field(s) that they know very little about to begin with.
Make no mistake. To be engrained in the tech sector is not a proxy for expertise in the underlying domain. It’s a common conflation and folly.
Society and technology have been locked in an infinity loop since the stone age. We invent. It affects us. Which effects what we invent. Which affects us. But our greatest inventions tend to be overlooked. There’s a case to be made for a good many breakthroughs, and recency bias does drift to communications, and to zeroes and ones and now, to AI. But for my money, the single greatest triumph of the modern era is birth control: revolutionising the quality of life for women everywhere, in some cases saving their lives, especially in the global south. Its effect rippled – nay, roared – into commerce, politics, careers, and families.
Tech like that changes everything. Existentially. So, are we in the same postcode? Is the proclaimed forthcoming bot apocalypse about to reshape our collective trajectory?
The marketing riddle
Well, if bots simply scan the market and make transactional decisions focussed on primary utility, what does that mean for mental availability – ala, brand? Is it still important? Or, by virtue of such a paradigm, does physical availability, once its partner in crime, now trump it entirely?
The loudest agentic commentators won’t answer those questions. Rarely, have they even heard of the terms. But while they don’t know it, this is the very thing they claim is being disrupted.
And what of Double Jeopardy?
For the uninitiated, the law(s) of double jeopardy, discovered by William McPhee in the 60s and generalised to brand theory by Andrew Ehrenberg in the 70s, state that smaller brands are disadvantaged in both rates of growth and of customer loyalty. That’s the double, and they’ve been validated repeatedly ever since.
Thus, it seems a reasonable suspicion that in a bot world, this would exacerbate as bigger brands can afford to establish more sophisticated, expensive, programmatic-style physical availability. In other words: designing content for machine consumption. This means that bots are more likely to calibrate in favour of larger firms not dissimilar to search and share of voice, where they already dominate.
But is this really going to happen? Well, partly yes, and partly no.
One factor is how the bot is trained, and by whom. In one extreme, if we all have such a de facto bot-decision-maker, we’d best prepare for a kind of dystopian ‘samesville’.
Everyone turns up to the office wearing the same uniform loafers, colour matched to the exact same beige jumpsuit, rocking with the same unisex bowl cut, carrying the scent of the exact same toiletries, slightly offset by the whiff of tuna found in the same lunchbox.
“Morning!”, grins a gregariously positive Truman Burbank to his neighbours, all teeth and charm. As his petri-dish of an existence is orchestrated all around him, he bathes in a life with no agency. “Oh, and in case I don’t see ya, good afternoon, good evening and good night!”
Of course, that might soften a little, perhaps with variations by category or dare I say it, in the hazy concept of a ‘segment’. For instance, all men under 40 with 3 children – and other select elements – buy a particular car, courtesy of the botty one.
Of course, this is where the agentic preachers jump back in, rushing to correct the hypothesis.
“You don’t understand”, they swagger. “People will be training their own personal bots. We’ll make them do what we want them to do. That scary extreme just wouldn’t happen, okay?”
But their argument just ate itself.
If we’re still doing the driving, perse, then mental availability – the foundation of brands since forever, remains firmly in place. Not a speck of disruption to brand repertoires and category preferences, especially in higher involvement categories. Humanity and commerce remain connected and even after 5000 years, dyadic trade is a long way from dead. Maybe, just maybe, we won’t have to marry our oldest child off to the leader of local bot village.
And then, consider the case of my bung toe.
Bung toes and messy identity
I had a quite active sporting career at school. Football, as in soccer, cricket both indoor and out, a decent dollop of basketball, plenty of swimming, land yachting with my Dad, and a mountain of rugby.
Early one year, I was training for the latter when I broke one of my small toes – quite badly too. Being determined not to miss the season, I strapped it up like a man with a Houdini fetish, pumped the pain killers, and played through. I must have re-broken it about 5 million times and now it looks like a set of bald quokka Siamese twins.
So, when I buy shoes, that little hunchback can be quite painful if they’re too tight. Compounding that, I find that sizes can vary massively between manufacturers and continents. I mean, if we can land on the moon, you’d think we could sort out universal shoe sizes.
Anyway, that means I have shoes in my closet that range from a modest size 7 – no, shoe size is not related to anything else, right up to size 10 – or maybe it is. Then the shape of the shoe itself, and its materials, are a factor. The point is… good luck training the bot on that!
So, whether it’s the calamity of my little toe, deviations by manufacturers and product, category ingress and egress, or the inevitable fluctuations in personal choice, there are big holes in the consumer bot bucket, dear Liza.
And still, the plot doth thicken.
My MBA students learn about Identity Theory. Why? Because it turns out that our identity is informed by what we do, and that what we do is informed by our identity. It’s kind of a circular phenomenon, which is complicated by the fact that we hold different identities at the same time.
A man can believe himself to be an authoritative leader, dutiful son, passionate lover, and devoted father – simultaneously. Well, it turns out that the way we perceive and value products, and interactions, can be affected by a perceived identity, or combination, triggered by external unknown factors. This example, along with many others, reminds us that humanity is not so easily conquered by code.
Smoke and mirrors and smoke
It’s true that when tech commentators make predictions about markets, customers, economics, and human behaviour that, well…. they’re tech commentators.
Smoke. And it’s true that in the marketing and customer domains over the last couple of decades, the tech tail has been busy wagging the dog.
Mirrors. But sometimes, where there is smoke, there’s actual fire – and that’s the case here. We may see working use cases, certainly in low involvement categories. And there’s no doubt that AI is changing the face of search. But that’s all part of physical availability. An evolution for sure, but a revolution?
Not remotely. Inevitably, this will create more smoke though, because these things tend to get assessed in isolation, and simplistically, usually by those without a wider understanding of the connected tissue of market and customer economics.
As it stands, many consumer businesses struggle to deliver reliability and coherency across existing customer interactions. That’s not because it can’t be done, far from it, but many of the popular marketing technology use cases are only popular due the afore mentioned smoke. While this doesn’t materially affect growth itself, it sure is corrosive to profit.
Bolting on bots before solving for that, will only exasperate the statistics, and AI will merely speed the errors.
So, when you encounter the next breathless botty hallucination, just remember that it’s always better if the curtains match the carpet – and for now at least, a bot doesn’t get that reference. I guess we should be thankful for that.
