Why Access Isn’t Enough – The AI Adoption Gap is More Than Technology
Not too long ago (Like NRF 2024) the question in retail technology used to be whether AI was real. That debate is over. AI adoption is accelerating at a 23% CAGR. Retailers now direct an average of 15% of their IT budgets to AI-specific initiatives, and year-over-year AI spending is rising by 27%.
The money is moving. The software is being purchased. And yet, as my IHL colleague Jerry Sheldon shared this week, the million-dollar question isn’t about access anymore. It’s about whether those tools are actually changing how work gets done.
That gap, between having the tools and truly embedding them into daily workflows, is the defining challenge of 2026 retail technology. This week’s case studies make it unmistakably clear.
Macy’s $640 Million Lesson
Macy’s is in the middle of its “Bold New Chapter” turnaround, and the company is not being timid about capital commitment. Its China Grove fulfillment hub in North Carolina represents a $640 million investment in a single, highly automated robotic facility designed to handle approximately 30% of Macy’s entire digital supply chain capacity.
The goal is to double productivity in fulfillment operations. AI tools are being deployed for demand planning and forecasting. SAS Customer Intelligence 360 is being used to drive more personalized customer engagement.
By any conventional measure, this is what “going all in on technology” looks like. And yet Macy’s own COO Tom Edwards described the company’s generative AI journey as “very early innings.”
That admission deserves attention. Macy’s operations leader is telling the market they’re barely getting started on the AI layer that will ultimately determine whether those investments create durable competitive advantage. This is the back-office version of the adoption gap. You can spend hundreds of millions on robots and cloud systems. Getting AI truly woven into the fabric of daily decision-making takes longer and requires different skills than deploying hardware.
IHL’s own research confirms the upside when retailers do close this gap: AI-powered demand forecasting reduces forecast error by 20% to 40%. AI-driven markdown optimization improves sell-through 10% to 20% while preserving 2 to 4 points of margin. AI personalization engines increase conversion 3% to 7% and average order value 5% to 12%.
Macy’s is building the infrastructure for these outcomes. The hard work is the integration and the culture change. Which in both cases is TBD.
The Frontline Challenge: Technology Adoption Is More Than A Technical Problem
Move from the warehouse to the store floor and the adoption gap becomes even more visible and more contentious.
Electronic shelf labels are a useful data point. Retailers like Walmart and Kroger are deploying them at scale, seeing a clear path toward pricing agility and operational efficiency.
Yet the UFCW, the country’s largest grocery union and one of the most powerful unions overall, launched a national campaign this week to ban them, citing concerns about job displacement and what they call “surveillance pricing”. It looks like the COVID risk that allowed retailers to install these units is no longer a concern large enough to override employment risks.
Whatever your perspective on the policy debate, the UFCW campaign is a reminder that technology adoption in retail isn’t purely a technical or financial problem. It’s a human and political one. To be fair, surveillance pricing is a stretch when it comes to individual users, but it is totally accurate when it comes to spying on the prices of competitors and making those changes in the field. The UCFW’s choice of language in their challenge is a good choice of words that can be construed in many different ways.
The Tesco example runs in the opposite direction. Over 1,200 Tesco convenience stores in the UK are now equipping their employees with body cameras, a technology rollout motivated entirely by worker safety rather than productivity optimization or cost reduction. Tesco even built its own proprietary system to manage footage securely, and the program has been extended to home delivery drivers.
These two stories together illustrate something important. Technology deployment in retail is not monolithic. Some of it is about margin. Some of it is about safety. Some of it generates workforce friction. Some of it reduces it. The retailers who succeed are the ones who lead with honesty about which is which, invest in change management alongside technology deployment, and measure outcomes that matter to their people as much as outcomes that matter to their P&L.
This Week’s Implementations: The Scale Is Staggering
Any lingering doubt about the pace of AI adoption should be dispelled by the sheer volume of implementations announced in a single week.
In supply chain and back office operations: US Foods is using AI to convert handwritten notes into digital orders. Associated Wholesale Grocers is deploying Relay AI across forecasting for 3,500 stores. Heineken is running a generative AI system they call Cobra to connect 150 of their breweries. The scale of that last one alone is worth sitting with: a generative AI network spanning 150 manufacturing facilities across a global enterprise.
On the customer-facing side: 7-Eleven is expanding its frictionless checkout pilot. White Castle has deployed an AI-powered security guard that automatically resolves 91% of its security incidents. Amazon continues adding capabilities to Rufus, its AI shopping assistant.
At the strategic partnership level: Loblaw launched a grocery shopping app directly inside ChatGPT, giving Canadians the ability to shop conversationally through an AI interface they already use. Unilever signed a five-year deal with Google Cloud to build what the company calls an “AI-first digital backbone.” Wesfarmers committed to enterprise-wide agentic AI deployment across Kmart, Officeworks, and Priceline through a multi-year Google Cloud partnership.
IHL data shows that profit winners are 224% more likely to prioritize embedding generative AI into existing applications. Loblaw’s ChatGPT integration and Wesfarmers’ agentic AI commitment are textbook executions of that strategy.
The Real Barrier Is Not Awareness
The data on where retailers stand is sobering. According to IHL Group’s study of 300+ mid-market retailers, 71% cite ROI uncertainty and 71% cite integration complexity as the top barriers to technology transformation. Only 49% believe their technology investments deliver ROI within 12 to 18 months.
Only 25% of retail organizations report strong in-house technical expertise. More than half of IT budgets remain committed to maintaining existing systems, leaving just 45% available for the advanced technologies that could drive competitive advantage.
Jerry Sheldon’s prescription, from his analysis this week, is practical and actionable: embed large language models into recurring workflows rather than standing up separate AI tools employees have to remember to use, and start with the tasks people actively dislike doing. Those are the highest-engagement AI testbeds.
The insight is correct. The retailers closing the adoption gap are not the ones with the most AI licenses. They’re the ones who have made AI the path of least resistance for daily work as a first step. Then rethinking processes totally with AI in the center for the future as people get more comfortable with the tools.
The momentum is undeniable. AI is going to reshape how work gets done in retail, not someday, but now, at speed and scale. The retailers making moves this week, from Macy’s warehouse robots to Heineken’s brewery AI network to Loblaw’s ChatGPT storefront, are not running experiments. They’re making infrastructure decisions that will define their competitive position for the next decade.
Retailers who are still “evaluating” while their competitors are lapping them on implementation are not just leaving operational efficiency on the table. They are ceding the capability to compete on the terms that will define retail for the next decade.
Close the gap. The tools are there. The only question is whether the workflows are changing.
Listen and Learn More
Read all 40+ articles here at https://www.ihlservices.com/newsletter/
For AI-powered retail intelligence built on 30 years of IHL Group proprietary research: https://www.justaskgreg.ai
About the Author
Greg Buzek is President of IHL Group, a retail and hospitality technology research and advisory firm with more than 30 years of proprietary data on retail transformation. He is the host of the Retail Reality Check podcast and the creator of JustAskGreg.ai.
Related Research
• IHL Shelf Intelligence Report: https://www.ihlservices.com