On Wednesday evening, Red Badger hosted a fantastic dinner at L'Escargot in Soho, partnering with Dynamic Yield and Amplience to bring together a small group of senior retail leaders to discuss how we create greater personal connection with our customers and the mechanics to make it happen.
I've spent most of my career inside retailers, watching brilliant, well-intentioned teams work incredibly hard to understand their customers. And the paradox is consistent: the data gets richer every year, the ambition is genuine, and yet the experience a customer actually receives rarely reflects any of it. The personalisation feels bolted on. The content is the same for everyone. The best customers are greeted like strangers.
That gap between what retailers know and what they show was the thread running through the entire evening.
The View From Outside Your Own Four Walls
Before the broader discussion opened up, Robbie Grant from the Mastercard Economics Institute offered a perspective that reframed the room's thinking — and it was a useful place to start.
Most retailers are working from an incomplete picture of their customers. The data they hold reflects what happens inside their own ecosystem: what a customer bought, browsed, returned, redeemed. But that's only one chapter of a much longer story. Robbie's work at the Mastercard Economics Institute sits at the intersection of transaction intelligence and consumer behaviour at scale — and what that vantage point reveals is striking. Where customers are spending outside your walls. Which life events — a move, a new job, a growing family — are quietly shifting their priorities. Where loyalty that looked solid is already beginning to erode, often to a competitor you wouldn't have predicted.
The implication for personalisation is significant. If the signals you're acting on only reflect what customers do with you, you're optimising for a partial relationship. The moments that matter most to a customer — the ones that shift their needs and open them up to a brand that gets it — often happen somewhere you can't see. Transaction intelligence changes that. It extends the view beyond the edges of your own data, and surfaces patterns that make the difference between a timely, relevant experience and another generic one.
It was a timely reminder that knowing your customer isn't just about depth of data within your own platform. It's about breadth — and the humility to recognise that the customer in front of you has a life that extends well beyond your last interaction with them.
From Adobe to Amplience: What Sam O'Neill Learned Leading John Lewis's Content Transformation
The first thing Sam O'Neill said that made the room sit up was this: the migration from AEM to Amplience at John Lewis wasn't really a technology decision. It was a capability decision.
Sam led the onsite experience programme at John Lewis & Partners, with accountability spanning content, ad tech, analytics implementation, design systems and web architecture — the full stack of what it takes to deliver a coherent customer experience at scale. And she'd arrived at a clear diagnosis: the content infrastructure John Lewis had wasn't built to serve individuals. It was built to serve channels.
AEM, used well, is a powerful tool. But for a retailer with the ambition to make every customer feel genuinely known — not just segmented — it had become a constraint. Content lived in silos. Updates were slow. Reuse was manual. The gap between "we know what this customer wants" and "we can show them something that reflects it" was a structural problem, not a prioritisation one.
The move to Amplience, and the COPE (Create Once, Publish Everywhere) architecture that came with it, was the foundation for something more fundamental: the ability to think in components rather than pages. To create content once and have it adapt — by channel, by segment, by moment — without a new project each time.
What Sam was honest about, and what resonated with everyone in the room, was how much of this work is organisational before it's technical. Migrating millions of assets is a known problem with known solutions. Shifting how a content team thinks — from "what's going on the homepage this week" to "what's the right content for this customer, right now, on any surface" — is a different kind of change. It requires different sign-offs, different briefs, different definitions of done.
If you needed to run genuinely different content experiences for twenty different customer segments tomorrow, could your current setup do it? Not in a pilot. In production.
For most of the people at the table, the honest answer was no. Where has the needle moved at John Lewis? Sam was clear that progress is real — automation is unlocking scale that manual workflows never could. But she was equally clear that the work is ongoing. The platform is only as useful as the organisation's willingness to use it differently.
Building a Testing Culture at Primark: What Annette Rowson Found When She Arrived
Primark is an unusual business to lead experimentation in. A four-year-old ecommerce offering. Huge footfall. A brand built on value and discovery rather than recommendation and retention. And — as Annette Rowson was refreshingly candid about — a data culture that, when she joined, was still finding its feet.
Annette is Analytics and Experimentation Lead at Primark, she knows what it looks like to land in an organisation that says it values data and then makes decisions on instinct. She also knows what it takes to change that and it's rarely what people expect.
The tools aren't the hard part. Standing up an experimentation platform, defining a testing framework, building a dashboard — those are tractable problems. The harder work is cultural. It's convincing a senior stakeholder that a test which doesn't confirm their hypothesis is still a good outcome. It's building enough psychological safety that teams are willing to be wrong in a structured way.
What Annette has built at Primark is a genuine test-and-learn culture, the kind where experimentation is embedded in how decisions get made. She spoke about the commercial case she had to construct internally: not just "experimentation is best practice" but "here is the revenue we left on the table by not testing this."
The question her contribution surfaced for the room was one that landed with some weight: are we using data to genuinely learn, or are we using it to confirm? Most organisations, if they're honest, are closer to the former.
What the Wider Conversation Revealed
The conversation naturally took an organic turn which perpetuated questions that made people pause before answering.
When asked where customer data lives, the answers were illuminating. It lives in multiple systems, owned by multiple teams, requiring significant orchestration to do useful things. The data exists at volume. Acting on it coherently is a different problem entirely.
In most organisations Personalisation is not a golden thread that runs through the strategy. Often teams are operating in siloes, marketing, content, product and customer insight, creating friction and fractured experiences for the customer. What was discussed was the need for outcome-led thinking and shared OKRs to frame the strategy for personalisation from a customer perspective and deliver against their connection to the brand.
Asking our dinner guests to think with their consumer hats on; some of the best personalised experiences were retailers where data was being used in a pragmatic early funnel aspect creating resonance with buyers and engendering conversion. Trinny London was famed for the ability to match skin preferences to product types. Suit Supply for the size passport, meaning your measurements follow you as a customer from store to site and Nike, whose size recommendation based on what other customers have bought, helps reduce returns rates and actually caters for size fluctuations giving customers more confidence to buy.
The evening didn't produce a tidy conclusion, and it wasn't meant to. What it produced was a clearer picture of where the real barriers are, and they're not where most roadmaps focus.
The technology exists. The data exists. The intent exists. What's harder is the organisational alignment, the platform decisions made years ago that now constrain what's possible, and the gap between personalisation as a strategy and personalisation as something a customer actually experiences.
That's the gap worth closing. And the first step, as the evening made clear, is being honest about where you actually are.