Only 5% of retailers get AI right amidst lack of integrated systems

The remaining 95% fail because when executives discuss AI, the answers are less clear.

Only five percent of retailers that invest in artificial intelligence (AI) succeed because they embed it directly into operational processes and require it to deliver measurable results.

Alex Luu, regional director of Philippines and Vietnam at Centric Software, said that for these companies, the starting point is not the algorithm but the operational structure. “AI depends on an end-to-end process where data, people, and systems operate on the same platform, and when these elements are disconnected, it cannot function effectively,” he said during the Retail Asia Summit 2026 held at Shangri-La, Makati.

According to a report from MIT, the remaining 95% fail because when executives' discussions move to specifics—what type of AI, which use case to prioritise, and how it connects to operations—the answers are less clear.

This disconnect becomes more pronounced as businesses expand. “This challenges how quickly information from different sources can be collected and turned into decisions,” Luu said.

He said speed matters as inventory decisions have direct financial consequences. One example cited during his presentation involved markdowns. In 2024, Nike reportedly marked down about 45% of its items to clear inventory.

“Allocation decisions also depend on this level of insight,” Luu added. For example, a store in Bonifacio Global City may require a different assortment than a store in Makati.

The same principle applies when retailers operate dozens or hundreds of stores across the Philippines or expand to other markets such as Thailand, Vietnam, or Malaysia.

Luu said that AI can support these decisions, but only if the underlying operational structure allows data to move quickly between teams and systems.

“For retailers considering AI, the technology itself is not the starting point—it must be an operational system where data, processes, and teams are connected.”

Once that structure is in place, Luu said that AI can analyse data, support faster decisions, and help retailers respond to changes in the market.