Most UK online retailers say their AI projects are still failing short, even as shoppers grow more used to AI. New research commissioned by Quickfire Digital, based on 2,005 UK consumers and 201 online retailers with ten or more employees, finds that 77 per cent of retailers see areas where their AI initiatives need work, while 27 per cent of consumers say they have not had any positive AI experiences at all.
The same study estimates that UK ecommerce retailers invested about 120 million pounds in AI technologies in 2024, with around 92 million at risk of underperformance and the exposure projected to reach 230 million pounds by 2030. Yet 37 per cent of shoppers say they have abandoned purchases because of complicated checkouts, 19 per cent dislike being forced to use a chatbot, and 54 per cent cite high delivery fees as the main deterrent.
The research also suggests that some applications show value for customers. Around one third of respondents say they appreciate automated returns and refund systems, 32 per cent point to AI-generated product descriptions or images, and 31 per cent highlight AI-powered customer support, while retailers most often label AI chatbots, data analysis tools, AI-driven marketing and content generation as underperforming.
Other online platforms face similar pressures to keep their technology aligned with user expectations, especially when activity depends on reliable information. This is evident in sectors where player behaviour is closely monitored, and many users prefer the option to play legally at UK sites as those services rely on analytics and customer insight to maintain consistent performance.
For UK ecommerce retailers, that same need to keep their systems working in ways customers find reliable now coincides with a market growing more slowly than in previous years, so AI budgets face closer scrutiny. Recent figures support this, and according to the Ecommerce Delivery Benchmark Report, 84 per cent of ecommerce businesses reported growth in 2024 while the average rate was 5.6 per cent, pointing to a market where gains are becoming harder to secure. Slower growth leaves retailers less room for error in AI spending.
Quickfire Digital cofounder Martin Harper says retailers now need to concentrate on execution rather than experimentation. He points to four priorities: using AI to improve experience and retention, making loyalty programmes more effective and personalised, giving service teams better tools and analytics and reinforcing checkout, payment and trust fundamentals.
Retailers that want AI to play a bigger role in service are being pushed to show that automation can handle more than routine questions. Research on AI-based chatbots in customer service and their effects on repurchase and trust reports that efficiency gains from chatbots can be undermined when they fail on complex issues, which in turn weakens satisfaction and future purchase intent. That concern is reflected in the Quickfire findings, where retailers place service quality, deeper analytics and stronger checkout basics among their main priorities for improving AI performance.
Most UK online retailers acknowledge gaps in their current AI tools, while consumers still encounter checkouts, fees and service experiences that make them reconsider a purchase. At the same time, 23 per cent of retailers in the study say all of their AI initiatives have been successful, indicating that when investment focuses on removing real friction rather than adding more features, AI can still provide an advantage in a tougher ecommerce climate.
