Rory Sutherland: Why marketing’s biggest risk in 2026 is mistaking efficiency for progress

Advertising and innovation, he mentioned, sit uncomfortably outdoors that worldview. And AI, for all the joy round it, dangers being pulled firmly into the identical cost-obsessed logic.

“The invoice for AI goes to come back due,” Sutherland mentioned. “Insane quantities of cash have been spent. So how do you promote it to shoppers? The best manner is price discount.”

To clarify the hazard, he returned to certainly one of his favourite analogies, what he calls the “doorman fallacy.” Strip a doorman’s job all the way down to opening a door, substitute them with an computerized system and have fun the financial savings. What disappears from the spreadsheet is all the pieces else the doorman did: safety, recognition, reassurance, standing.

“You stroll away not being held remotely liable for any of the worth destroyed,” Sutherland mentioned.

He sees the identical sample repeating throughout fashionable enterprise, usually inspired by consultants engaged on gain-share agreements. They will declare credit score for lowering prices, he argued, with out being held accountable for misplaced worth, misplaced gross sales or long-term injury.

Grocery store self-checkouts, in his view, are a textbook instance. As an possibility, they make sense. As a default, they quietly erode the expertise.

“Finance notices it’s cheaper to get clients to do the work, so it goes from being an choice to kind of an obligation.”

The result’s shoplifting, frustration and techniques that work in concept however collapse in actual life. “It’s inconceivable to do a big household store if it’s important to test it out your self,” Sutherland mentioned. “Self-scanning would work brilliantly if people had three arms.”

On the root of this, he argued, is a deeper philosophical divide about what companies are for. One custom, related to economists like Peter Drucker and the Austrian faculty, sees advertising and innovation as central to worth creation. One other, rooted in shareholder worth considering, treats enterprise primarily as an effectivity machine.

“The aim of a enterprise is to search out and maintain a buyer,” Sutherland mentioned, citing Drucker. “There are solely two features that add worth: advertising and innovation. Every little thing else is a price.”

That concept, he famous, now sounds nearly radical. Effectivity considering dominates as a result of it suits neatly into spreadsheets. Subjective worth doesn’t.

“Advertising contributes perceived worth. However perceived worth is tough to quantify. And what’s laborious to quantify will get ignored.”

AI, he warned, dangers being absorbed into that very same bias. In its first section, he expects it to be offered and imposed as a headcount-reduction software, very similar to self-checkout was. “The primary section is similar however cheaper.”

What worries him most is how simply companies undervalue human interplay when it resists measurement. In service industries particularly, he argued, the human component disproportionately shapes satisfaction.

He supplied an instance from Royal Mail. Years of funding in bettering supply reliability had no measurable affect on how a lot folks appreciated the model. What did matter, it turned out, was whether or not folks appreciated their postie.

“In case your postman was a little bit of a bastard, you didn’t like Royal Mail,” Sutherland mentioned.

That hole between what issues to clients and what companies measure creates what he known as a “quantification bias.” Quick metrics crowd out gradual ones. Belief, retention and lifelong worth lose out to short-term effectivity beneficial properties.

This bias additionally explains why corporations push clients into their lowest-cost channels. Cellphone numbers disappear from web sites. Self-service turns into obligatory.

“A customer to the location converts at about 0.5%,” Sutherland mentioned, citing an internet journey enterprise. “Anybody who telephones us converts at about 30%. Are we actually positive that is clever?”

Trying forward, Sutherland sees three doable phases. The primary is the present rush towards automation and cost-cutting. The second, he hopes, will see some companies utilizing AI to enhance experiences relatively than strip them again.

“We are going to see such a profusion of automation that there’s a real alternative to do the alternative. Completely different is healthier than higher.”

That change, he steered, could also be simpler for founder-led or family-owned companies, which aren’t trapped by quarterly reporting. Public corporations, he mentioned, usually battle to maintain severe advertising as a result of short-term monetary stress crowds all the pieces else out.

The third section is extra radical. Drawing an analogy with the electrification of factories, Sutherland mentioned new know-how solely delivers actual beneficial properties as soon as processes are redesigned round it.

“In the event you can produce actually good content material rapidly and cheaply, do you have to simply proceed doing what you used to do, or do you have to reinvent the method altogether?”

He predicts a future the place companies work extra proactively, producing work and not using a transient after which discovering patrons for it. Cannes, he joked, would possibly turn into much less of a retrospective and extra of a commerce truthful. “We’ve bought these things. Do you need to purchase it?”

Sutherland closed with a bit of recommendation he described as unintentional however basic. “In the event you’re a marketer, don’t promote what you do. Promote the way you suppose.”

The actual worth of selling, he suggests, is perspective. Engineers and finance groups optimize what they’ll measure. Entrepreneurs flip the issue round and ask the way it feels to the human on the opposite facet. “Until you will have a marketer within the room, you’re prone to doing issues which might be severely fucking dumb.”