Michael Scheibner is CEO and Chairman of the Board at GK Software, a Fujitsu company and global leader in retail technology.
Retail leaders are serious about implementing AI. According to Gartner, 91% of retail IT leaders have prioritized AI as the top technology to implement in 2026.
There’s a good reason for this, as AI can enhance personalization and optimize pricing, promotions, fraud detection and real-time associate guidance at the point of transaction across convenience stores, fuel stations, grocery outlets and other online and brick-and-mortar retail formats.
This momentum is also being driven by changes in consumer behavior. A recent Adobe report found that U.S. e-commerce site visits originating from generative AI tools increased 393% year over year in March 2026 and are of a higher quality than other traffic from other channels.
There’s a problem, though: Many retailers lack the right IT infrastructure and connected data to take advantage of AI. Retailers are aware of this issue, with data infrastructure being the top priority for AI adoption, according to a recent survey by Verizon Business and Cisco.
Disconnected systems and data are barriers to AI.
Retailers have typically taken a “best-of-breed” approach to implementing technology over the years, layering new applications on top of old ones, connecting them with complex APIs and interfaces in order to deploy new services.
When it comes to AI adoption, this IT infrastructure can become increasingly difficult to maintain.
This approach has enabled retailers to quickly deploy new services, but it also creates highly complex, fragmented technology infrastructures. For instance, siloed, disconnected data can lead to inaccurate inventory information. This, in turn, can lead to several other issues: out-of-stocks, lost sales, inconsistent pricing and other disconnects.
The financial impact can be high. Out-of-stocks alone cost retailers $82 billion in lost sales in 2021, according to NielsenIQ research.
Disparate systems and inaccurate data also affect other areas of retail, including the store, which depends on consistent, shared information for accurate pricing, personalized promotions, inventory availability and much more.
As noted by research firm IHL Group, retailers are having “a collective ‘aha’ moment about AI: it’s completely useless if your data is a mess. Bad data just gets you faster bad results.” Their research found that “the retailers experiencing the highest sales growth rate prioritize data cleaning, training and storing at rates 110% higher than sales laggards.”
That’s why a unified architecture—combining all sales channels and front-end/back-end into a single architecture—can play a role in AI success. It can help ensure that the same sales data, customer profiles, loyalty information, inventory location and availability and other essential data is continually updated and available in real time through both physical and digital channels.
Transitioning To A Unified Architecture
Unified retail is an architectural transformation, and making the transition involves several distinct stages.
Prior to implementation, retailers must do a reality check. As organizations map out their strategy, they must first define how the data will flow in their unified architecture.
Then they must align all stakeholders around shared KPIs. For example, measuring customer lifetime revenue as a key metric, rather than revenue by channel (since unified commerce is designed to unite sales channels).
Finally, they must audit current systems and decide what to keep or replace.
During the actual implementation, organizations must ensure that data synchronization—inventory, pricing and customer data, for example—is consistent and in real time across all touch points. They must also make sure not to disrupt operations as new workflows go live.
To reach these goals, retailers can focus on composable architecture (characterized by API-first, modular systems); establish cross-functional teams made up of IT, business and operations team members; and prioritize high-impact cases where unified commerce delivers a clear ROI.
Finally, post-implementation, retailers should focus on testing new strategies to deliver consistent experiences. One area to focus on is building unified customer profiles by bringing together data from multiple sources, as this is a crucial factor for delivering personalization, which is a key benefit of AI adoption.
Unified systems can help accelerate AI adoption.
2026 can be the year that AI begins to deliver on the promise of retail transformation, and unified systems and data are an important foundation.
Retailers want to know that their AI investments will be equipped for the long haul. Starting with the right foundation is a key aspect of adapting to a quickly changing AI environment. When implemented strategically, a unified data architecture can help ensure that retailers reap the benefits of an AI-driven future.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
