Sebastian Roland, Group Head of Strategy, Impact BBDOAI is everywhere. In brainstorms, on whiteboards, inside briefs. But while everyone’s trying to figure out what to do with AI, the more important question is this: What data are you feeding it?
AI isn’t the differentiator anymore. The tools are available to everyone. The prompts are getting sharper. The outputs are improving. And yet, most brands are still scraping the surface of what’s possible; not because of the technology, but in no small part because of the inputs.
The recent ChatGPT 5 launch, widely seen as underwhelming, prompted The New Yorker to ask: “What if AI doesn’t get much better than this?”. But this provocation may miss the mark.
While model capabilities will likely continue to improve as compute power increases, the plateau of AI’s real-world impact is more likely to stem from our inability to feed these systems with truly meaningful and novel data. Perhaps the better question is: “What if our training data doesn’t get much better than this?”
The quiet problem: We underutilise what we need most – data
Here’s the tension that doesn’t get talked about enough: most brands don’t fully own or understand their own data.
A Forrester study found that only 32 per cent of marketers are confident in their organisation’s ability to control and activate their first-party data. And it shows. Many brands are still relying on fragmented dashboards, disjointed platforms, and siloed systems that can’t talk to each other.
At the same time, segmentation has stalled. With incomplete datasets, many teams default to broad-brush targeting and demographics over dynamics. A Deloitte study found that less than half of surveyed consumer felt brands were delivering on personally relevant marketing. Campaigns that feel increasingly generic in a world that demands more for creativity to be effective and impactful.
The deprecation of third-party cookies and tightening privacy laws have made the situation even more urgent. First-party data isn’t just a nice-to-have anymore. It’s a creative necessity.
From performance fuel to creative IP
Here’s the opportunity: to stop thinking of data as something that lives in spreadsheets and start seeing it as something that can shape stories.
AI offers brands a chance to do more than optimise. It offers a chance to unlock the creative and strategic potential of the data they already own, but haven’t yet realised the value of.
Start thinking about data not as a performance lever, but as your intellectual property.
And not all data is customer data. Nor does the most valuable data always sit in the marketing department. Some of the richest, most human sources of insight are hiding in plain sight.
Imagine:
- A fashion retailer’s archive of fabric swatches – a tactile timeline of shifting taste and trend.
- A bank’s record of small business loans by region – a behavioural map of ambition and resilience.
- A food delivery app’s unfulfilled orders – a digital diary of cravings and micro-desires.
- An airline’s passenger feedback to cabin crew – a 35,000-foot view of gratitude, empathy, and cultural nuance.
- A QSR’s real-time menu customisations – a flavour fingerprint of identity across rituals and regions.
This kind of data isn’t just goes far beyond the informational and goes in to the emotional connection people have with your brand, product or service.
In your data lie narratives that can and should be brand-defining. And when it becomes the source code for what an AI model sees, learns from, and builds with, it gives the machine something real to work with. Something only your brand can offer.
What gets in the way of using data
So why aren’t more brands doing this?
Because data still lives in silos. A study by Dataversity found that 68 per cent of data specialists said siloing of organisational data was the biggest barrier to innovation in areas such as AI. Because it’s often treated as a hygiene factor rather than a strategic asset. Because many organisations are set up to collect, not connect.
There’s also a mindset shift required. One that reframes internal knowledge from operational residue to cultural capital. One that treats qualitative archives and messy spreadsheets not as problems to clean up, but as material to mine.
Brands that succeed in the next wave of AI-powered creativity won’t just be those with the best ideas. They’ll be the ones who know their past. Who understand their customer journeys not just as funnels, but as stories. And who use AI not to synthetically replicate what already exists but go think far bigger to reflect something only they could say.
So what should leaders do?
Start with the basics. Audit what you actually have. Don’t just look at your CRM or analytics dashboards but look at your archives, your product logs, your customer service transcripts, your R&D notes.
You’re probably sitting on more original, emotionally rich data than you think.
Then: connect it. Structure it. Protect it. And crucially, use it as the springboard for creativity that solves the biggest and most interesting business problems in your organisation.
Therein, lies the real potential of the AI revolution we find ourselves in.
Final thoughts
In the generative era, every brand has access to the same models. What they don’t have is your data. Your nuance. Your fingerprints.
Simply put, it’s time to get passionate about turning your brand’s data in to your advantage.
The next chapter of marketing won’t be won by those who prompt better. It’ll be won by those who feed better. Because in the end, the most compelling creative won’t come from AI.
It’ll come from brands that know who they are, and know how to turn what they know into what the world feels.
By Sebastian Roland, Group Head of Strategy, Impact BBDO
