Data and AI: small words, big consequences
Retail leaders are absolutely sure they need to invest in AI because the opportunities appear to be unlimited but the quandary for many is that they have not yet seen any return on their current investments and so are uncertain about where exactly to commit further money.
Needless to say, surveys abound when it comes to AI as organisations seek to find intelligence to help them decide where to place their bets. Calling AI investment decisions bets seems to denigrate its importance but it might not be too far from the truth for many companies who are being sold all aspects of AI very heavily.
Subscribe to TRBAgentic AI is certainly hot right now. According to Logicbroker over a third of leaders say AI will drive half of all transactions by 2027 and that 90% expect AI agents to influence at least 20% of online orders within this tight time-frame. This might have been a finding Logicbroker wanted as it describes itself as an agentic commerce orchestration engine.
Whichever way you cut the data CEOs in retail and other sectors are worried about how to deal with AI in all its myriad incarnations. The biggest question on their minds is whether they are transforming fast enough to keep pace with technological change, with 42% of leaders citing this as their top concern, according to the PwC 2026 Global CEO Survey. This has undoubtedly contributed to them lacking much confidence in delivering revenue growth this year, with only 30% expressing belief in generating growth in 2026.
It’s not through a lack of investing in AI though, it’s the disappointing returns on their investments to date where they have been suffering. Only 12% of CEOs say AI has delivered both cost and revenue benefits. For many of these executives they have been typically piloting AI on a piecemeal basis, says PwC.
Where there has been much greater success is with companies that have deployed AI at scale across the whole organisation and laid solid foundations on which AI can be successfully implemented. Just as with e-commerce, success is determined by the quality of the plumbing and the movement of data through these pipes.
All too often AI is being layered onto fragmented systems and inconsistent datasets. Customer data is often split across e-commerce platforms, in-store systems, and loyalty schemes, making it difficult for AI to form a complete and accurate view of the customer. Even the most advanced AI tools don’t have the intelligence needed to deliver meaningful results.
This situation has been recognised by Sam Jayes, head of client services at Sagacity, who says: “Retail’s AI race is accelerating, but for many brands, meaningful returns still aren’t materialising. While plenty of retailers are piloting AI initiatives, very few are successfully scaling them. More often than not, the barrier is the quality and integration of the data that underpins these investments.”
She adds: “In a market where margins are tight and competition is intense, the gap between retailers who get data right and those who don’t is widening. The winners won’t be those who adopt AI fastest, but those who fix their data first.”
The situation today echoes to some extent the transformative period of the early 1990s when executives understood the internet would dramatically change their businesses and life generally. We know today that many retailers have continued to place insufficient effort into organising their data and the failings of this resonate powerfully. There is a risk that many are rolling on their lack of a handle on their data from the e-commerce world into that of AI and the problems will be multiple-times more consequential.
