Rory Sutherland warns: AI risks repeating Google's mistakes - 7 lessons for marketers
NewsThursday, 09 July 2026 at 20:28
Marketing thinker Rory Sutherland says artificial intelligence is at a crossroads. According to Ogilvy’s vice chairman, AI risks falling into the same commercial trap as Google, where ad revenue ultimately outweighs information quality. In a wide-ranging conversation on AI, marketing, and behavioral psychology, Sutherland shares insights that go far beyond branding.
For Sutherland, the real question isn’t which AI model performs best today—it’s which business model will end up shaping the answers users receive.
1. “Ads aren’t the problem. The incentives behind them are.”
Sutherland stresses he’s not opposed to advertising in principle. His concern is what happens once an AI platform becomes dependent on ad revenue.
He points to Google’s evolution. Early on, search results were relevant, with clearly labeled ads alongside them. As commercial pressure grew, the experience shifted. Paid results took over more space, and the most relevant information no longer surfaced by default.
His warning is blunt: the same dynamic could play out with AI assistants if ads become the dominant revenue stream.
2. “Bad actors can outspend good companies”
Sutherland argues that ad auctions don’t reward relevance—they reward profitability.
That creates a perverse incentive. Competitors, intermediaries, or even shady operators often have more budget to buy visibility than the original, high-quality source.
If AI answers become commercially skewed, he says, public trust in these systems could be fundamentally damaged.
3. “AI is being sold mainly as a way to fire people”
Sutherland also questions how companies are positioning AI right now.
Too often the emphasis is on cost-cutting and headcount reduction, with far less attention on innovation or creating new value.
He warns this mindset can trigger a negative spiral. If organizations use AI primarily to eliminate jobs, it can sap employee motivation and erode social acceptance of AI.
4. “Platforms are turning into protection rackets”
One of his sharpest critiques targets Big Tech’s power.
Sutherland likens the trajectory of online advertising to a protection racket. Ads once helped brands get discovered; now, he says, companies must pay just to remain visible at all.
He cites Amazon, Facebook, and LinkedIn as platforms businesses have grown increasingly dependent on.
His key takeaway: never build a company that relies entirely on a single digital platform.
5. Fear of rejection beats rational arguments
Beyond AI, Sutherland shares a classic marketing lesson from his time at American Express.
For years, the company tried to persuade consumers to apply for a credit card with rational arguments. The real barrier, however, was psychological.
Many interested people wanted the card but hesitated to apply for fear of being rejected.
A subtle shift in messaging—from “apply for a card” to “we’d love you to join”—largely removed that barrier.
According to Sutherland, effective marketing often means removing emotional friction rather than piling on more reasons.
6. Innovation starts by asking: “What are we still doing this for?”
Sutherland also addresses innovation inside established companies.
His advice is strikingly simple. Organizations should regularly ask themselves:
- Which activities no longer add real value?
- What problem are we actually solving today?
- What psychological barrier are we missing?
- Which audience are we still failing to reach?
He says questions like these often drive more innovation than elaborate brainstorms or new management frameworks.
7. Don’t just ask why customers bought—ask why they almost didn’t
His final insight focuses on customer research.
Most organizations analyze only successful purchases. As a result, Sutherland says, they miss crucial signals about doubt and psychological resistance.
He cites a company that asked customers one extra question right after purchase:
“What was the one thing that almost stopped you from buying this product?”
That question, he says, uncovers insights that can significantly improve future conversion.
Why these lessons matter now
The conversation suggests AI’s toughest challenges may not be technical, but economic and psychological.
As companies race to build ever more powerful language models, the real battleground is who decides what information users see—and which commercial incentives sit behind it.
At the same time, Sutherland argues successful AI adoption isn’t just about tech. Organizations that understand human behavior, motivation, and decision-making will gain more from AI than those focused solely on automation and cost-cutting.
His analysis aligns with a broader industry debate: the future of AI will be shaped not only by model quality, but by the incentives designed by the companies that build them.
Latest comments
Loading