Earlier this year my wife and I were gripped by Apple Cider Vinegar, the Netflix drama based on Belle Gibson’s rise and fall. Gibson built a wellness empire on the claim she’d cured her terminal cancer through food and natural remedies, claims that turned out to be completely fabricated.
What struck me wasn’t just the deception. It was how major institutions, Apple, Penguin, and countless followers, got swept up in the story. Because it was a good one. Inspiring, persuasive, and emotionally compelling.
And it made me reflect: this isn’t just a wellness story gone wrong. It’s a mirror for how businesses operate.
When narrative wins over numbers
In marketing, we often find ourselves chasing stories that fit what we want to believe. A campaign worked because it “felt” right. A platform is scaled because the narrative is good, not because the numbers back it up.
That’s where the danger lies. Measurement becomes a supporting act, not the main event. We accept flattering data points, dismiss the rest, and create feedback loops that reinforce belief rather than challenge it.
I’ve seen this play out particularly in brand vs. performance debates. The narrative of opposing forces, long-term storytelling versus short-term sales, is compelling. But if we step back, the goal isn’t brand or performance. It’s profit.
Without a unified measurement framework, teams default to isolated KPIs. A high ROI might be praised, even if it reflects underinvestment. A brand campaign might be dismissed, even if it improves mental availability and long-term acquisition costs.
That’s not strategy. That’s theatre.
And now, AI has entered the chat
The stakes are only getting higher with AI. There’s a lot of buzz around intelligent tools promising automated insight and predictive targeting. But what happens when we can’t explain how those insights were generated? Or when the models are trained on biased data? Or worse, when they reinforce flawed metrics?
AI has the potential to scale great decisions. But it also has the potential to scale bad ones. And when we skip scrutiny in favour of storytelling, we risk falling into the same trap that enabled Gibson’s rise, believing the myth because we want it to be true.
From observation to cation: A toolkit for marketers
So how do we avoid building our marketing strategy on shaky ground? Here are five ways marketers can balance insight with integrity:
– Build a unified measurement framework: Don’t silo measurement by function, format, or channel. Bring together brand, performance, media, and CX teams under one umbrella with a shared definition of success. A clear north star, whether it’s customer lifetime value, margin efficiency, or brand equity, ensures everyone is pulling in the same direction. This reduces contradictory KPIs and helps avoid optimisation in isolation.
– Pressure test the data: When a result is convenient, interrogate it. Always ask: what would have happened if we did nothing? What other factors could be influencing this result? Encourage a culture of constructive challenge, where “why not?” is as valued as “why yes.” This mindset prevents vanity metrics from becoming decision-making metrics.
– Watch for AI snake oil: AI solutions often come with slick demos and bold promises, but not all are fit for your business. Insist on transparency in how models work and what data they’re built on. Evaluate whether outputs are explainable to non-technical stakeholders. And remember, an algorithm trained on flawed data will only amplify those flaws. If you wouldn’t trust the source, don’t trust the system.
– Train for interpretation, not just insight: It’s not enough to know what a dashboard says, you need to know what it means. Invest in training that improves data literacy across teams, not just in analytics functions. Run workshops where teams simulate decision-making based on ambiguous or conflicting data. The goal isn’t just competence, it’s confidence.
– Embed governance from day one: Data and AI governance can’t be retrofitted. Define from the outset who is accountable for metrics, how data is validated, and how decisions are documented. Implement QA rituals across teams and make bias identification part of the process. Just like financial controls, measurement governance needs oversight, escalation routes, and periodic review.
Closing thought: Are we asking the right questions?
The most dangerous stories are the ones we don’t question. Belle Gibson’s downfall came not just from her deception, but from the silence of those who should have asked more.
The same applies to our work.
If we’re going to scale AI, launch brand campaigns, or invest in new platforms, let’s make sure we’re not just telling ourselves what we want to hear.
Let’s measure what truly matters.
James Heimers is SVP of analytics at Rapp UK
