Find out how artificial intelligence is helping convenience retailers to reduce crime, save time and grow sales.
By Sarah Britton
With so much talk about deepfakes, the death of creativity and the threat to jobs, it’s understandable that retailers might be wary of this powerful and fast‑evolving technology. But there’s no denying AI’s ability to collate and analyse data from multiple sources in record time, joining up the dots before most of us even put pen to paper.
Convenience wholesalers are taking the plunge and embracing AI head‑on. JW Filshill and Parfetts have just teamed up with Wholepal, an AI‑powered platform that automates the processing of new‑line forms, improving accuracy and saving time.
Meanwhile, wholesale buying group Unitas has launched an AI Academy in partnership with educational platform PAIR, for members and supplier colleagues, with a view to seeking out AI opportunities to future‑proof their businesses.
But what about retailers themselves – how can they harness AI to improve the day‑to‑day running of their business?
“I think most retailers are aware of AI but maybe don’t know how to use it quite yet,” says Anand Cheema, owner of Costcutter Fresh in Falkirk.
“We’ve had it for a long time. We used to work with Veesion with our cameras about four years ago now and then we installed Retail AI last year and took Veesion out. We’re always trying to evolve.”
Watch and learn
Veesion is an AI video‑surveillance software system that analyses CCTV footage and can recognise suspicious gestures. Retailers then get real‑time video alerts when shoplifting or questionable behaviours are detected.
RetailAI offers a multi‑pronged approach. Its anti‑theft module can integrate with CCTV, picking up on theft‑related behaviours and sending video alerts to staff. It can also trigger a siren and warning message via in‑store radio telling shoplifters they’re being watched.
In addition, the radio, which is also through RetailAI, can play targeted in‑store promotions and adverts, as well as music.
“The radio is amazing,” says Anand. “It’s personalised and you can sell that space to suppliers so you can quickly get your money back on the initial outlay and then the AI function of the radio as well.
“A customer could be walking down the confectionery aisle, which would automatically trigger the AI to start advertising chocolate brands like Cadbury or Galaxy, or sweets on promotion with your symbol group.”
SLR #ThinkSmart Innovation Award winner, Girish Jeeva, is always first in line for new technology and has recently been trialling RetailAI’s new SmartWatch, which is synced to the firm’s anti‑theft module and can instantly send staff videos of suspected shoplifting incidents, as well as aiding team communications and task delegation.
He has also adopted smart queue management, which detects queues and makes an audio announcement to redirect shoppers. “AI helps guide customers to self‑checkout when normal tills are busy, improving flow and reducing wait times,” Girish explains.
Last year, he installed Innovative Technology’s MyCheckr — an AI‑powered age‑estimation and facial‑recognition device — at his beer‑cave entrance. The door automatically opens if someone appears over 25, but if the tech deems a customer to be underage, the door remains closed and customers have to show staff ID to gain access. “This reduces staff checks and speeds up service, especially at busy periods,” he says. “We don’t get any shoplifting [in the cave] at all since we installed it.”
Deciphering data
Another way retailers are using AI is for data analysis. “I haven’t got AI integrated into my EPoS, but I use it separately,” says Anand, who has signed up to the premium versions of Google NotebookLM and ChatGPT. “I’ve been tinkering about with it for the last eight or nine months.”
He recently used the AI platforms to help generate suggested orders of Perfect Draft at Christmas. “I took all the sales data from the last four years, put it into AI and it chucked up a spreadsheet for me of a predicted order for this year.
“You have to feed it information, but it reads the sales data, so it sees the decline, it sees the increase at Christmas, it looks at customer trends, dates and times as well. It looks at what day Christmas falls on and the stock availability. I used to do it all manually myself with predicted orders, but it’s just so time consuming. Using AI easily saves me a full day’s work.”
Asif Ashraf, who owns four stores in central Scotland, and also owns EPoS provider MHouse, is keen to tap into AI to make running a c‑store easier. Like Anand, Asif also used to upload a lot of reports onto ChatGPT and gauge insights. “I’ve always been a big data person, but it’s difficult to do a comparison across my stores every single week,” he says.
He is currently running a beta version of the firm’s MPOS solution, which incorporates AI and is due to launch later this year. “Once a week, our current system autogenerates a health‑check report, which is emailed to the owner of the store,” he says. “The beta version compares all four of my stores and flags up what it sees as issues, opportunities and weaknesses.”
Reports such as this are available on various EPoS systems, but the time retailers need to analyse them isn’t nearly as easy to come by. The beauty of AI‑led EPoS is that it can also make educated assertions as to what the data implies. “It highlights products that we should think about delisting,” says Asif. “For example, in one of the stores we had two different sizes of fresh chicken breasts. It said over the last three months you’ve had quite a high amount of wastage, which has mitigated any profit we’ve had on this item. So it suggested that, if we delist one, we’ve still got a different size of chicken breast and it will allow that to be sold at a better price point.”
It will also give insights across products, pricing and availability and highlight any issues on ‘hero’ products such as bread and milk.
The beta system can also monitor how often different staff members are using voids and overrides at the till. “It flags up every pattern it sees, for you to pay attention to,” says Asif.
“It gives me action plans for all my store managers too, stating what to do over the next week or month with promotional plans.”
If you aren’t particularly tech‑savvy, then one of the easiest ways you can use ChatGPT to improve your category performance is with a photo, claims retailer and MHouse owner, Asif Ashraf.
“In the past I’ve uploaded an image of my soft drinks fixture. I’ve told ChatGPT: ‘I’m a convenience store. I’m based on the main road. If you were a retail expert for the retail multiples, what changes would you make to this display? What about my range – what do you think I could do?’”
The system acknowledged that he had merchandised the fixture based on the trends towards energy drinks and health but flagged up that he had too many Monster and water varieties and highlighted how delisting some lines would make room for two extra shelves of value carbonates.
The personal touch
Asif believes AI will enable EPoS to function “on a more human level”. “At the moment, suggested ordering is based on existing stock levels and average sales etc. This [new version] will be from more of a human point of view.
“It will have access to what promotions your symbol group is running, and it will up those quantities because it should recognise that these things will sell out more because they’re on half price.”
The AI system will also recognise when it’s needed. “It may be that you’ve forgotten to do your fresh food order on a Tuesday morning,” says Asif. “It’ll pop up and say: ‘You’ve only got 15 minutes left; I’ve already done the order for you. Do you want to skim through it and change anything? Or do you want me just to just transmit it?’ That type of automation is the dream we’re working towards on the MPOS side.”
He thinks that a more intuitive system could help retailers to make big savings. “From a retailer’s point of view, that’s exactly what I would need to mitigate the effect of rising costs that I’ve got across my stores and some of it’s quite crippling like staffing costs, electricity costs and rising food prices. It’s tougher than it used to be and unless we adapt and find a way then we probably won’t be here after a few years. The ones that adapt are the ones that will survive.”
While Asif, Anand and Girish are ploughing ever further into the seemingly endless field of AI, there is still resistance in some parts of the sector. Anand believes older retailers are more closed‑minded when it comes to AI. “People are stuck in their ways or they don’t have the capacity to pick up on new technology as quickly as the younger generation,” he says. “And investment is a huge barrier as well.”
But with costs continuing to spiral and AI offering numerous money‑ and time‑saving tools, can the convenience channel really afford to bury its head in the sand?
Girish thinks not. “AI and tech haven’t been gimmicks — they are integral to transformation: raising sales, improving security, enabling new services, and increasing efficiency.”
If you run a small convenience store (like many independents) and the idea of “AI” sounds big or intimidating — here’s a step‑by‑step, manageable approach.
- Start small with a clearly defined pain point — e.g. theft, underage alcohol sales, queueing at peak times, or demand forecasting. Choose one area where improvement will make a noticeable difference.
- Choose turnkey/managed solutions — many AI systems for small retailers are now offered as SaaS (software as a service), or managed services. You don’t need an in‑house IT team; the provider can install, maintain and support the system.
- Pilot before full rollout — try the system on a trial basis (e.g. one camera zone, or age‑verification at one door, or self‑checkout lane) to test how it works, how staff and customers react, and whether ROI is acceptable.
- Train staff and communicate with customers — make sure staff know how to use the system; make customers aware of new tech (for example via signage) so they’re comfortable. Transparency helps build trust and avoid backlash.
- Monitor results and iterate — track key metrics (sales, losses, theft incidents, stock‑outs, customer feedback). If things improve, expand; if not — re‑evaluate or pause.
- Prioritise privacy, fairness and compliance — especially when dealing with CCTV, age‑verification or behavioural tracking. Be mindful of data protection, local laws/regulations and customer sentiment.
- Use technology to complement, not replace, human service — maintain friendly, community‑oriented service; treat AI as “assistive” rather than “human‑replacing”. This helps preserve a local shop feel while reaping efficiency benefits.
