Related Articles
Speaking at The Drum’s Predictions 2026, the vice-chairman of Ogilvy UK warned that businesses are sleepwalking into a future defined by cost-cutting and regulatory fear. AI, he argued, is being sold in exactly the wrong way and marketers will pay the price if they don’t push back.
Rory Sutherland did not open his Predictions session with optimism. “My predictions for 2026 are slightly pessimistic,” he said, before outlining what he sees as the dominant mindset across much of business.
“Most businesses only have two principal modes,” he argued. “Cost reduction and regulatory paranoia. Don’t fall foul of the regulator. Don’t have a reputational disaster. Spend as little money as you can.”
Marketing and innovation, he said, sit uncomfortably outside that worldview. And AI, for all the excitement around it, risks being pulled firmly into the same cost-obsessed logic.
“The bill for AI is going to come due,” Sutherland said. “Insane amounts of money have been spent. So how do you sell it to clients? The easiest way is cost reduction.”
To explain the danger, he returned to one of his favorite analogies, what he calls the “doorman fallacy.” Strip a doorman’s job down to opening a door, replace them with an automatic system and celebrate the savings. What disappears from the spreadsheet is everything else the doorman did: security, recognition, reassurance, status.
“You walk away not being held remotely responsible for any of the value destroyed,” Sutherland said.
He sees the same pattern repeating across modern business, often encouraged by consultants working on gain-share agreements. They can claim credit for reducing costs, he argued, without being held accountable for lost value, lost sales or long-term damage.
Supermarket self-checkouts, in his view, are a textbook example. As an option, they make sense. As a default, they quietly erode the experience.
“Finance notices it’s cheaper to get customers to do the work, so it goes from being an option to more or less an obligation.”
The result is shoplifting, frustration and systems that work in theory but collapse in real life. “It’s impossible to do a large family shop if you have to check it out yourself,” Sutherland said. “Self-scanning would work brilliantly if humans had three hands.”
At the root of this, he argued, is a deeper philosophical divide about what businesses are for. One tradition, associated with economists like Peter Drucker and the Austrian school, sees marketing and innovation as central to value creation. Another, rooted in shareholder value thinking, treats business primarily as an efficiency machine.
“The purpose of a business is to find and keep a customer,” Sutherland said, citing Drucker. “There are only two functions that add value: marketing and innovation. Everything else is a cost.”
Want to go deeper? Ask The Drum
That idea, he noted, now sounds almost radical. Efficiency thinking dominates because it fits neatly into spreadsheets. Subjective value does not.
“Marketing contributes perceived value. But perceived value is hard to quantify. And what’s hard to quantify gets ignored.”
AI, he warned, risks being absorbed into that same bias. In its first phase, he expects it to be sold and imposed as a headcount-reduction tool, much like self-checkout was. “The first phase is the same but cheaper.”
What worries him most is how easily businesses undervalue human interaction when it resists measurement. In service industries especially, he argued, the human element disproportionately shapes satisfaction.
He offered an example from Royal Mail. Years of investment in improving delivery reliability had no measurable impact on how much people liked the brand. What did matter, it turned out, was whether people liked their postie.
“If your postman was a bit of a bastard, you didn’t like Royal Mail,” Sutherland said.
That gap between what matters to customers and what businesses measure creates what he called a “quantification bias.” Fast metrics crowd out slow ones. Trust, retention and lifetime value lose out to short-term efficiency gains.
This bias also explains why companies push customers into their lowest-cost channels. Phone numbers disappear from websites. Self-service becomes mandatory.
“A visitor to the site converts at about 0.5%,” Sutherland said, citing an online travel business. “Anyone who phones us converts at about 30%. Are we really sure this is intelligent?”
Looking ahead, Sutherland sees three possible phases. The first is the current rush toward automation and cost-cutting. The second, he hopes, will see some businesses using AI to improve experiences rather than strip them back.
“We will see such a profusion of automation that there’s a genuine opportunity to do the opposite. Different is better than better.”
That change, he suggested, may be easier for founder-led or family-owned businesses, which are not trapped by quarterly reporting. Public companies, he said, often struggle to sustain serious marketing because short-term financial pressure crowds everything else out.
The third phase is more radical. Drawing an analogy with the electrification of factories, Sutherland said new technology only delivers real gains once processes are redesigned around it.
“If you can produce really good content quickly and cheaply, should you just continue doing what you used to do, or should you reinvent the process altogether?”
He predicts a future where agencies work more proactively, producing work without a brief and then finding buyers for it. Cannes, he joked, might become less of a retrospective and more of a trade fair. “We’ve got this stuff. Do you want to buy it?”
Sutherland closed with a piece of advice he described as accidental but fundamental. “If you’re a marketer, don’t sell what you do. Sell how you think.”
The real value of marketing, he suggests, is perspective. Engineers and finance teams optimize what they can measure. Marketers flip the problem around and ask how it feels to the human on the other side. “Unless you have a marketer in the room, you’re at risk of doing things that are seriously fucking dumb.”
Source link
