AI in retail. It’s the shiny new toy every marketer is desperate to flaunt.
Today, algorithms and predictive analytics promise to ‘revolutionise’ how we shop. But here’s the thing: for all the fanfare, many retailers are still struggling to get the basics right. Empty shelves? Clunky layouts? No AI in the world can fix that if you don’t have your act together first.
So before we hand over the keys to the geeks, let’s take stock of where AI is genuinely making a difference. Not in the hype, but in the practical, day-to-day mechanics of retail.
Eye in the sky
In 2024 Morrisons rolled out AI-powered cameras across its stores, promising a brave new world of shelf replenishment and inventory nirvana. These digital eyes continuously monitor conditions, capturing images to track stock levels, flag low-stock items, and expose empty shelves.
AI is transforming the in-store retail experience across the UK by improving operational efficiency and customer engagement. For retailers, a clear use-case for AI lies in its ability to anticipate demand and prevent gaps before they happen
And in an age where empty shelves can sink customer loyalty faster than you can say “click and collect,” that’s no small thing. Morrisons reported in their September trading update, a +2% improvement in shelf availability since the introduction. Sometimes, the future of retail is simply about ensuring the shelves are stocked today.
Your special offer is personal
In May 2024, Tesco introduced ‘Clubcard Challenges,’ an AI-driven initiative offering tailored tasks to Clubcard members, enabling them to earn up to £50 in points over a six-week period. These challenges are customised based on individual shopping habits
Additionally, Tesco plans to expand its use of AI to further personalize customer interactions. By analysing data from its extensive Clubcard user base, Tesco aims to provide tailored suggestions and insights, helping customers make healthier choices, reduce waste, and save money. Ken Murphy, Tesco CEO speaking at FT Future of Retail Summit in September 2024 remarked “I can see it nudging you over time, saying: ‘I’ve noticed over time in your shopping basket that your sodium salt content is 250% of your daily recommended allowance. I would recommend you substitute this, this and this’,”
AI is turning grocery shopping into something smarter, as and when the time and attention is paid to getting it right
When help becomes intrusive
AI has delivered impressive wins in retail like smarter shelves, personalised offers, and streamlined operations. Also true to say that not every innovation hits the mark.
Remember Clippy, Microsoft’s overenthusiastic paperclip assistant? Always there, rarely actually helpful ! Many retailers are now deploying AI-driven virtual shopping assistants to ‘help us’ to shop online. They risk treading a similar path. Designed to enhance the online experience with personalised recommendations and real-time support, they can just as easily become a nuisance. Pushing products you don’t want or need, popping up at the wrong moment, or misreading your preferences—it’s a delicate balance between assistance and annoyance. If not executed thoughtfully, AI helpers risk alienating shoppers rather than aiding them. Convenience is key, but there’s a fine line between being helpful and being, well, Clippy 2.0
As researchers at DVJ, we have always followed the data. The trick, we maintain, is to use a variety of approaches because that helps provide real insight
Ask, Observe and Understand the Shopper
Shoppers are more than data points—they’re people with needs, habits, and quirks. Truly understanding them demands more than numbers; it requires a blend of empathy and sharp observation.
Data tells us what shoppers did and it’s the purest form of shopper measurement. At DVJ, we always start with behavioural data whenever available. Sales data lays the foundation: what’s flying off the shelves, what’s gathering dust, and how promotions are driving (or not driving) incremental sales.
Still, without understanding the why behind those numbers, it’s like staring through a rear-view mirror.
That’s where observation steps in. Simulated research environments such as virtual shelf tests let us see behaviour in action. In these retail-realistic setups, we watch shoppers navigate the aisle, interact with products, and make decisions. Do they linger or move on? Pick up an item or glance and ignore it? These small but critical details reveal the subconscious cues that drive decisions, free from memory gaps or self-justifications.
Simulated environments also unlock deeper insights with advanced tools. Tachistoscope testing (T-Scope), for example, captures the instant impact of visual stimuli, showing products for mere milliseconds to reveal what grabs attention before conscious thought kicks in. Search-and-find tasks add more perspective, challenging shoppers to locate specific products on a virtual shelf. This uncovers how easily they identify items, revealing the effectiveness of packaging, shelf placement, and category organisation.
Once the behavioural puzzle is pieced together, the art of asking begins. Not endless surveys, but focused, meaningful questions that dig deeper. A well-placed story-cue can uncover likes, dislikes, or the real needs a category is serving.
Turning Insights Into Action
Of course, the role of Insight is to help brands and retailers stay close to what the shopper really needs. As the possibilities for AI increase, well-structured research will guide us on what to do next and be led by the shopper. For retailers and brands, it’s about translating facts into strategies that work on the shelf and in the shopper’s mind. Data, observation, and robust questioning together help us to find the truth. ‘In God we trust, all others must bring data’
Take shelf layouts, for example. By combining sales data with behavioural observations, retailers can fine-tune category flows, placing high-velocity products where they’re most visible and grouping items to align with natural shopper missions.
Communication, too, benefits from this ask-and-observe approach. Knowing what draws attention in-store helps brands design packaging that pops, signage that speaks directly to shopper needs, and promotions that feel less generic and more personal.
It’s the researchers’ role to apply the tools and thread together the evidence.
