November 27, 2025
The grocery and FMCG industry has spent the past decade wrestling with digital transformation — but as we head towards the end of 2025, things feel different. At Grocery Gazette Horizons, a new event dedicated to the sector’s evolving challenges, one theme dominated every session, panel, and side conversation: AI is no longer experimental. It’s operational — but unevenly so.
Across retailers, producers, and technology partners, the story was clear. Demand for AI-enabled efficiency, accuracy and personalisation is accelerating, yet legacy systems, fragmented data, and trust barriers are slowing progress. I had the opportunity to join a panel alongside leaders from Hanshow and Olio to explore where AI can deliver real commercial value today — and where retailers still need to catch up.
Below are the key insights I took from the event, and what they mean for grocery in 2026 and beyond.
Katie Hunter, General Manager at Uber Grocery and Retail UK, highlighted a challenge everyone in the room recognised: retailers all speak different data languages. Something as simple — and critical — as allergy information varies in naming conventions, formats and structures.
When an ecosystem depends on accurate data flowing between grocers, marketplaces and logistics partners, this lack of standardisation becomes a structural blocker. AI thrives on clean, consistent inputs, and too many retailers are feeding models with data that doesn’t align.
The businesses that solve interoperability will move fastest.
Trigo’s approach to autonomous stores stood out for its pragmatism. Rather than reinvent infrastructure, they’re able to use existing CCTV to build a digital twin of every aisle. This allows them to:
AI adoption doesn’t always require moonshots — sometimes it’s about creatively re-using what’s already there.
Sarah Anderson, Global Head of Category Analytics at Mars, framed elements of their AI strategy around a simple question: “Is it going to make life easier?”
It’s a refreshing counterpoint to the tech-first rhetoric often dominating AI conversations. From store associates to supply planners, the businesses that design for augmenting humans — not replacing them — will build trust faster, scale initiatives more easily and achieve more meaningful change.
One of the key themes from our panel was the limitations of traditional ERPs in forecasting. They look backwards and struggle to incorporate external data such as:
Modern AI models, by contrast, operate at a far more granular level — store-by-store, SKU-by-SKU. This shift enables more dynamic pricing, smarter replenishment, and improved commercial decision-making.
We explored the commercial upside of AI-driven price changes, particularly for perishables. Instead of end-of-life markdowns, gradual/phased reductions can reduce waste and improve margin.
But when it comes to personalisation, retailers must avoid what was identified as the “ick bell curve” — where helpful becomes intrusive. Customers want relevance, not creepiness; and they want to know they’re getting the best price (dynamic pricing at an individual level isn’t the silver bullet some think it is).
The trust gap is real:
Retailers must bridge this by being transparent, responsible and consistent.
Across the day, one trend kept resurfacing: while AI understanding is increasing, legacy technology is preventing many organisations from progressing at the right pace.
Some grocers are deploying digital twins, dynamic pricing engines, and operational AI today. Others are still constrained by rigid ERPs, unreliable data flows, and siloed architectures. Unless retailers accelerate modernisation, this gap will continue to grow — and the competitive consequences will become more visible.
Before implementing AI, ensure foundational data problems are addressed. AI amplifies both strengths and weaknesses, and although AI can assist with data standardisation, it’s important to fix the data, not obfuscate inaccuracies.
Modular modernisation can deliver value in the short-term without the need for wholesale ERP replacement.
If colleagues (at all levels of the organisation) don’t trust or understand the solution, it won’t scale.
The line between helpful and uncomfortable is thin. Customers will reward clarity; content is key (e.g., using AI to interpret why a customer might be calling a Contact Centre, personalising the response without the need for endless keypad pressing to find the right person).
Digital twins, micro-forecasting, and dynamic pricing are all tangible entry points.
If you’re exploring how to reduce waste, improve forecasting accuracy, or simplify store operations using AI, I’d be happy to share more from the panel discussion and dive deeper into where retailers are already seeing results.
Get in touch — let’s turn AI from a strategic ambition into a practical advantage.
Paul Sims is a Retail Technology Strategy Consultant, who joined Equal Experts earlier this year after spending the last 20+ years working for retailers across the globe, including positions as CTO at Halfords and Chief Architect at New Look, Primark, Marks & Spencer, and Argos in Shanghai.
All business models have their pros and cons. But, when you consider the type of problems we help our clients to solve at Equal Experts, it’s worth thinking about the level of experience and the best consultancy approach to solve them.
If you’d like to find out more about working with us – get in touch. We’d love to hear from you.
