Why Unilever's All-In Influencer Bet is Wrong.

I began my marketing career at Unilever in Australia. After completing my marketing degree, it was at Unilever, long considered the "university of marketing", where my real education happened.

Standing in supermarket aisles, watching people shop, building end-cap displays, and having conversations with real customers taught me more about marketing than any lecture or textbook. These human moments were what marketing was fundamentally about.

So when I read now that Unilever's new CEO Fernando Fernandez is dramatically increasing their influencer marketing investment, from 30% to 50% of total marketing budget and expanding to "20 times more content creators", I felt a curious mix of emotions.

On one hand, doubling down on human connection makes intuitive sense. On the other, this massive shift toward influencers comes precisely when AI is transforming how brands build relationships with consumers. The timing couldn't feel more paradoxical.

The most troubling aspect of Fernandez's new strategy isn't just the lopsided investment in influencers—it's the fundamentally flawed premise behind it. During a 'fireside chat' with investment analysts, Fernandez referred to big brands as "by default suspicious" and suggested this inherent suspicion could be overcome by "creating marketing systems in which others can speak for your brand at scale."

This statement doesn't just misunderstand consumer psychology—it undermines the very concept of what a brand is and how it functions.

Brands exist precisely to reduce consumer suspicion and uncertainty. They are trust marks, signalling consistent quality and values that consumers can rely on. The world's most valuable brands—Apple, Amazon, Toyota, Coca-Cola—are often the largest ones precisely because they've earned consumer trust at scale.

Consumers aren't suspicious of brands because they're big. They become suspicious when brands behave in ways that erode trust: when they deliver inconsistent quality, when their actions contradict their stated values, or when they fail to evolve with changing consumer expectations.

The idea that bigness itself creates suspicion is not just wrong—it's dangerous for a company whose entire portfolio consists of billion-dollar global brands. It's particularly perplexing coming from the leader of a company that has built some of the world's most trusted consumer brands.

If Fernandez truly believes Unilever's brands face a trust deficit, the solution isn't to outsource that trust to paid influencers. It's to examine why these brands might have lost consumer confidence and address those fundamental issues directly.

Fernandez's strategy centres on what he calls the "4V model of social media": volume, variety, virality, and velocity. By dramatically expanding the universe of creators representing Unilever brands, they aim to create a "machine of demand creation" through peer-led advocacy rather than traditional advertising.

The math seems compelling on the surface. But there's a critical flaw in this approach: consumers aren't naive about influencer marketing. Research consistently shows growing skepticism toward paid influencer content.

This points to an ironic twist in Fernandez's logic: If consumers are "by default suspicious" of big brands, they're arguably even more suspicious of people who get paid to speak on behalf of those brands. Exchanging direct brand communication for influencer endorsements doesn't solve a trust problem, it potentially compounds it.

What's more, by spreading Unilever's voice across "20 times more" influencers, they risk diluting brand identity rather than strengthening it. When everyone is speaking for your brand, who actually controls what it stands for?

The contrast with P&G is striking. While Unilever pushes further into influencer territory, P&G continues increasing its spend on traditional advertising, considering "superior communication" a "key vector" in growing its brands. P&G understands that owning your brand voice, and making it distinctive and trustworthy, remains essential even in a fragmented media landscape.

The timing of Unilever's all-in influencer bet couldn't be more precarious given the seismic shifts occurring in digital engagement. Just this week, Meta announced its "digital companions" initiative for Facebook, Instagram, and WhatsApp. These AI personas, some voiced by celebrities like Kristen Bell and John Cena, represent Zuckerberg's vision for social media's future.

While currently facing scrutiny for content moderation issues, these companions signal a profound shift, from human-created influence to AI-generated engagement. They're just the beginning of what will become increasingly sophisticated synthetic influencers that can personalise at scale in ways human creators never could.

The collision course is set: human influencers with authentic voices but limited scale versus AI companions with infinite scale but manufactured authenticity.

What's particularly concerning about Unilever's strategy is that it appears to be looking backward rather than forward. The influencer marketing playbook has been running for nearly a decade now. Doubling down on it in 2025, when generative AI is transforming every aspect of consumer engagement, feels like investing in print advertising in 2008 as smartphones were reshaping media consumption.

There's another critical dimension missing from Unilever's approach: the growing importance of how brands appear in large language models (LLMs).

As consumers increasingly rely on AI assistants and large language models for purchase decisions, brands must optimise for how they appear in these systems. When a consumer asks ChatGPT or Claude for product recommendations, these AI systems draw on their training data to formulate responses. If your brand isn't properly represented in that training data, you'll be invisible to the "agentic buyers" of tomorrow.

This "LLM share of voice" is rapidly becoming as important as traditional media share of voice. When consumers delegate purchasing decisions to AI systems—"Find me the best moisturiser for sensitive skin" or "Order my usual household supplies"—the brands that have systematically optimised for these AI-mediated interactions will win disproportionately.

Unilever's massive influencer investment may generate short-term social media engagement but does little to position their brands for this AI-mediated future. Influencer content, while potentially valuable for training data, isn't structured in ways that typically enhance a brand's positioning in LLM responses.

Unilever's influencer push creates a strategic dilemma for marketers across the industry. Should they follow Unilever's human-centric approach, or lean more heavily into AI-driven engagement?

The true challenge isn't choosing between human influencers and AI systems, it's orchestrating them together effectively. Here's what's missing from Unilever's approach:

Rather than outsourcing trust to an army of influencers, Unilever would be better served by reestablishing its brands as direct trust marks in consumers' lives.

This requires a fundamental reexamination of how their brands create value and build relationships. The problem isn't that Unilever's brands are too big, it's that some may have lost relevance or consistency in an increasingly fragmented market.

The solution isn't to hide behind influencers but to strengthen what made these brands successful in the first place: consistent quality, meaningful innovation, and values that resonate with consumers. These fundamental brand attributes matter whether communication happens through traditional advertising, influencer partnerships, or AI-driven personalisation.

Having worked with both global brands and cutting-edge technology providers, I believe the winning approach isn't either/or but both/and:

The history of marketing teaches us that new technologies don't typically replace existing channels—they transform them. Television didn't kill radio; it forced radio to evolve. Social media didn't eliminate television; it changed how we consume video content.

Similarly, AI won't eliminate human influence but will fundamentally transform how it works. The brands that thrive won't be those making all-in bets on either humans or machines, but those orchestrating them together in service of deeper, more meaningful consumer connections.

For a company once considered marketing's greatest university, Unilever's current approach feels like studying last decade's curriculum while the future races ahead. The question isn't whether human creators matter—they absolutely do—but whether investing half your marketing budget in them at the expense of tomorrow's capabilities is the right long-term strategy.

Fernandez's misunderstanding of brand trust mechanics doesn't just risk wasting marketing dollars—it risks fundamentally misaligning Unilever's approach to building and maintaining brand equity in a rapidly changing landscape.

The most trusted brands of tomorrow won't be those that hide behind armies of paid advocates. They'll be the ones that boldly and directly demonstrate their value, whether through human creativity, technological innovation, or most likely a thoughtful integration of both.

Instead of accepting the flawed premise that big brands are inherently suspicious, Unilever should embrace its scale and heritage as strengths. Rather than trying to appear smaller through fragmented influencer voices, they should focus on making their size meaningful to consumers through innovations that only a company of Unilever's scale could deliver.

What's your brand's plan for balancing human and AI influence? How are you preparing for a world where consumers increasingly rely on AI systems to make or guide their purchasing decisions? And most importantly, are you building a brand that consumers trust directly, not just through the borrowed credibility of paid advocates?

Human or Machine? Yes.

Brandflow is written by Justin Billingsley, who has spent his career on all three sides of the industry's table: senior client, global agency leader, technology founder. First published 1 May 2025 in the Brandflow newsletter on LinkedIn.