AI is being used more and more in customer service in the retail world, and it is not always well received. While shoppers are open to the use of AI, they can lack trust in the technology and question where their data is being used. Luke Cuthbertson, Head of the CX Consulting Practice at Route 101, explores the implementation problem and the need for transparency.
It’s no secret that times are tough for the retail sector. Rising costs led to thousands of shops closing their doors for the final time in 2025, and the economic climate does not seem set to improve. To add to this challenge, consumer needs are still rising, with customers now expecting near instant support. Retailers are therefore facing a near impossible challenge when it comes to customer service: how to cut costs and improve efficiency whilst still meeting consumer expectations.
For many, AI seems like the obvious answer. However, as retailers race to deploy AI-powered customer service, hoping it can deliver high standards of customer service for much lower costs, many are making a critical mistake. By opting for speed over strategy and treating AI as a quick technology upgrade they risk losing consumer trust and doing more harm than good.
Shoppers are not inherently resistant to AI. In fact, many consumers are already comfortable using automated support for simple, transactional queries. However, there are valid concerns around transparency, accuracy and data privacy that can lead consumers to shun some AI-driven customer interactions. When those interactions feel impersonal, misleading, or it becomes impossible to speak to a human being, the risk is that trust is broken completely. This is the challenge that retailers need to overcome if they are to take advantage of the new technology successfully.
The implementation problem
The problem with AI in retail customer service is very rarely the technology itself. More often, the problem lies in the implementation. Too many businesses still view AI primarily as a cost-cutting device, focusing solely on reducing headcount or deflecting contact volumes, rather than improving the customer experience. This leads to rushed rollouts that don’t have a solid strategy behind them.
One of the most common failings is poor data readiness. AI agents are only as effective as the systems and information they can access – clean, connected, real-time data is vital for strong performance. This is particularly obvious in high-volume retail queries such as ‘Where is my order?’ enquiries. Customers expect immediate, accurate answers, but without the right data, AI an only give vague or generic responses. This can create significant frustration with customers, especially if it is not then easy to escalate the query further.
Retailers also often underestimate the operational complexity of AI implementation. AI is not a plug-and-play solution that can simply be switched on and left to run. To deploying AI agents successfully, businesses need to carefully rethink workflows, customer communications strategies and governance, as well as their customer service aims and strategies. There should be a well thought out strategy in place before the technology is deployed.
That said, retailers should also resist the temptation to pursue perfection. Successful AI adoption is often iterative. Often the best approach is to ‘test and learn’, gradually building internal confidence and refining systems over time. Better to implement AI on a small scale slowly and deliver sustainable results than attempt a large-scale transformation overnight and get things badly wrong.
External expertise is often key to avoiding costly mistakes. Many organisations lack internal experience deploying AI, and in customer-facing environments the risks of getting it wrong are high. Poorly executed AI deployments can damage brand loyalty far more quickly than they improve efficiency. In retail, where switching costs are low and competition is intense, even a single bad experience can push customers elsewhere. It’s therefore worth taking the time and getting the support needed to ensure AI use is successful.
Transparency is essential
Aside from initial deployment, the other key area where retailers are getting AI wrong is transparency. Consumers are often willing to engage with AI when they expect to. The issue tends to be when businesses attempt to disguise automation, if customers feel like they are being deceived, trust will be lost. Retailers therefore need to be upfront when customers are interacting with AI and make escalation to a human simple and quick. This enables customers to still feel in control and valued when dealing with automation.
That sense of control is increasingly important as AI becomes more conversational and embedded across customer journeys. The best AI experiences will not trap customers in endless automated loops or prioritise containment over resolution. Instead, they will combine speed and convenience with clear pathways to human support when needed.
Customer expectations will continue to evolve
Over the next three to five years, consumer expectations around AI-powered retail support will continue to evolve rapidly as the technology improves. Customers will increasingly expect 24/7 assistance, faster resolutions and more natural, conversational interactions. AI agents are also likely to become more deeply integrated with mainstream platforms and large language models, reshaping how consumers discover products and engage with brands altogether.
But as AI capabilities accelerate, the retailers that succeed will not necessarily be the ones using AI the most. Instead, it will be the organisations who recognise that customer trust is key and build that into the foundation of their AI adoption.
