Sorrell: AI’s revolution for WPP, Google and Meta is here

For more than three decades, Martin Sorrell built WPP into the world’s largest advertising holding company by mastering the economics of the big creative campaign. Today, from the helm of his newer, nimbler venture S4 Capital, he is delivering a stark verdict: that blueprint is obsolete.
Over the course of a wide-ranging conversation about AI, Sorrell, one of the industry’s most closely watched figures, sketches out a future in which the billable hour vanishes, creative headcounts shrink and algorithms dictate where the world’s US$1.2 trillion in media spend flows. The shift, he argues, is not incremental but foundational, touching every corner of the business from the studio to the boardroom.
The first casualty of AI, in Sorrell’s view, will be the industry’s traditional pricing model. “Many of the big holding companies like Publicis, WPP and Omnicom are getting 25 to 40 per cent of their revenue from traditional creative materials, like copywriting and visualisation,” he says. “The time required to generate these materials is being severely reduced, and clients are aware of this.”

What once took weeks and cost tens of thousands of dollars – factoring in planning, briefs and multiple rounds of amendments – can now be drafted, visualised and iterated in hours at a fraction of the cost. Yet many legacy firms continue to bill on those old time frames. Sorrell calls this disconnect unsustainable.
“The agency model has to switch from time and materials, to unit‑based, output‑based or subscription‑based pricing,” he insists. The human consequences will be equally momentous: “You’re going to see fewer strategists and fewer creative people … There will be fewer of them, more highly paid, sort of an upper funnel based on strategic and creative capability, but it’ll be slimmer than it’s been in the past.”
A second front opens in personalisation. Sorrell describes the coming wave as “the Netflix model on steroids”, a reference to the streaming platform’s ability to spin a single show into thousands of tailored promotional assets for different micro-audiences. “Historically, we would produce for a Squid Game or a Narcos campaign maybe a million or a million‑and‑a‑half assets,” he says. “We can now do many multiples of that.”
If true, the changes to the industry will be striking. Creative production will shrink in headcount, while new kinds of talent will be required to manage, curate and quality‑control an explosion of AI‑generated material.

But producing personalised content at scale is pointless, Sorrell notes, unless it reaches the right consumer, which brings him to media planning and buying.
This is a roughly US$1.2 trillion global business that he likens to investment management, yet it still runs on largely manual or semi-automated systems. Today, around two-thirds of that entire spend is concentrated on just four platforms: Google, Meta, Amazon and TikTok. Despite that concentration, budget allocation across channels remains far from fully automated.
Sorrell expects that to change rapidly. He foresees AI optimising every dollar across digital and non‑digital channels in real time. For brands in Asia, where mobile penetration and digital adoption are among the highest globally, the promise is sharper targeting and better return on investment. The corollary, however, is that marketers must learn to interrogate those algorithms, lest they hand over strategic control to opaque black boxes.
While customer‑facing applications grab the headlines, Sorrell is equally animated by AI’s potential to overhaul back‑office operations. He points to projects such as outside broadcasting, where cloud and GPU infrastructure could slash costs by 80 to 90 per cent, enabling production teams to operate remotely rather than on-site. These are not marketing‑specific savings, but they will impact agency profit-and-loss statements just the same.
The most radical shift, however, may be organisational. Sorrell calls it “the democratisation of knowledge”. While AI is forcing redundancies in some areas, it is also allowing more people within an organisation to access the same vital data, breaking down informational silos and accelerating collaboration.

He illustrates the point with an example from outside advertising: Jensen Huang at Nvidia, no less, who reportedly has 51 direct reports – a number far beyond the 12 or 13 that management consultants like McKinsey or BCG consider optimal. “How does Jensen do it?” Sorrell asks. “AI is made available to everybody in the organisation. That’s a democratisation of knowledge. You can flatten the organisational structure, and you see Meta and Mark Zuckerberg talking about it in the same way.”
The result, for those who manage the transition adeptly, is a leaner, more responsive enterprise with fewer middle-management layers.
No conversation with Sorrell is complete without a view seen through a global lens, and he is particularly fascinated by the “unique marketing ecosystem” of Hong Kong and mainland China, where Western tech giants coexist with formidable domestic platforms. He cautions against the simplistic view that Google, Amazon and Meta have no Chinese presence: “They do, and they’re called ‘outbound Chinese’, [which is when] Chinese multinationals like BYD or Huawei build their businesses internationally using the platforms in the West.”
Yet his real curiosity lies with the home-grown champions: Alibaba, Tencent, ByteDance and Douyin, each generating tens of billions of dollars in revenue.
In particular, Sorrell highlights their accessibility. “These are open systems available to everybody. So they’re not closed systems. It will be really interesting to see how that develops over time.”
As Sorrell’s reflections on China suggest, the pace and shape of AI’s assault on marketing will vary by region and platform – and not every company will manage to pivot successfully. But the overall vector is unmistakable: faster production, hyper-personalised content, algorithmic media buying, radical cost efficiencies and flatter hierarchies.