A widening gap between the promise of AI-driven ecommerce and the reality of legacy retail systems is leaving consumers frustrated and retailers unable to deliver improvements, according to new research commissioned by ecommerce agency Quickfire Digital.
The findings point to a growing structural pressure across the sector, where legacy tech and heavily customised ecommerce platforms are increasingly limiting the ability of businesses to implement change they know their customers want.
Drawing on surveys of both consumers and retailers, the UK-based study found that nearly half of consumers (45%) are frustrated with AI-powered ecommerce experiences, specifically citing issues with AI-powered search, AI recommendations feeling too generic and AI recommendations showing items that are out of stock.
A further 44% of consumers reported frustration with personalised experiences more broadly, including seeing the same recommendations repeatedly, being targeted with irrelevant or inaccurate product recommendations, and feeling that brands do not understand their preferences.
On the retailer side, over three quarters (84%) of respondents said planned improvements have been delayed or not delivered in the past year, with 40% citing budget pressures, 25% citing security requirements and 23% citing data fragmentation as key barriers.
More than a third (40%) of retailers also say they are now spending over 25% of their ecommerce budgets maintaining existing systems, leaving less capacity for innovation or customer experience improvements.
Meanwhile, a separate study published by Quickfire Digital earlier this year found that 52% of retailers say they are adopting further AI use to drive revenue, despite shoppers’ frustrations with the technology.
Martin Harper, Co Founder of Quickfire Digital, comments: “Right now, as retailers continue to invest heavily in AI and customer experience tools, most of the industry narrative we’re seeing is ‘AI personalisation = good’. But our data is actually challenging that and revealing something no one is talking about; mid-market and enterprise brands are stunting their own growth because they refuse to abandon their legacy or custom-built tech stacks. And if retailers’ existing systems cannot support the pace of change they need to keep up with consumer expectations, AI may instead be actively putting consumers off.
“It’s also interesting to see a significant amount of the ecommerce budget being spent simply to keep legacy and heavily customised platforms running. Retailers know they need to make meaningful improvements for their consumers, but this hidden ‘Growth Tax’ being paid by retailers who prioritise owning their infrastructure over commercial agility is trapping them. This is one of the biggest challenges facing retail businesses, because without the right infrastructure in place, even the best customer experience strategy and AI tools will struggle to deliver results.”
The research also reinforces the continued importance of the sector’s website, with 45% of consumers saying they prefer to discover products by searching directly on retailer websites. This compares to just 16% who prefer discovering products through social media and content creators, challenging common assumptions that social commerce or third-party platforms are replacing owned ecommerce environments.
“Despite all the industry noise around social commerce, marketplaces, AI shopping assistants, discovery on TikTok, Instagram, etc, consumers are still saying the retailer’s own website is the main and most trusted place they use to find and buy products. For retailers who’ve over-invested in social or marketplace channels at the expense of their own site, this insight is a real wake-up call,” Martin added.
