In this Retail Reality Check podcast episode, host Claire interviews IHL President Greg Buzek to argue that artificial intelligence in retail can only succeed on a foundation of clean, accurate data. The episode opens with a stark warning that retailers will feed AI agents erroneous inventory numbers, leading to rapid, costly mistakes, and cites IHL’s 2026 study estimating $1.7 trillion in annual losses from bad inventory and that 44 % of retailers lack real‑time stock visibility. Greg likens most retailers to being stuck in traffic while a few early adopters have already built the data highways needed for AI. He traces the gap back to 2018, noting that a minority of retailers invested heavily in IT, achieving centralized, high‑quality data that now powers generative AI for data cleansing. IHL research shows these “winning” retailers enjoy sales growth ten times higher and profit growth six times higher than peers. The conversation highlights Snowflake’s emphasis on accurate data, consistent metrics, and business context, with a case study of Italian department store Rinascente cutting report processing time by 80 % through near‑real‑time data refreshes. The episode then presents a 14‑move supply‑chain playbook, emphasizing RFID for 98‑99 % inventory accuracy (up from 85‑90 %) and citing IHL’s study linking RFID use to 57 % higher profit growth in 2023 and projected 88 % in 2024. Store‑level technologies are also covered: Orange’s rollout of Pricer electronic shelf labels (ESS) across France, with IHL ranking ESS as the top technology tied to sales growth (adopters see 75 % higher sales); and Colruyt Group’s extensive self‑checkout deployment, with IHL showing sales leaders 141 % more likely to be current on self‑checkout. Greg’s advice is to verify shelf‑level inventory before signing any AI contract. The episode concludes with a rapid‑fire “lightning round” of recent retail tech news, ranging from AI‑driven predictive operations, Oracle Fusion Cloud ERP at Whitbread, Omilia voice AI at Taco Bell, Ulta’s ChatGPT plugin, Target’s MyDay app, Instacart’s Caper smart trolleys, to robot deliveries and retail media networks. The host ends with a provocative question: “Can your team tell you right now what is on the shelf?” and urges listeners to attend the upcoming Retail Agent Con on November 17, where real‑world AI agent results will be showcased.
Chapters
Retailers are about to give AI agents inventory numbers that are wrong, and those agents will act on them at machine speed. IHL's 2026 study puts the cost of bad inventory at $1.7 trillion a year, and 44% of retailers still can't see their stock in real time. Hi, I'm Claire, your host. Thanks for joining this week's Retail Reality Check. A podcast on the latest tech installs in retail and consumer product goods. We provide hard data so you can make smart decisions based on IHL Group research. Before we dig in, put November 17th on your calendar. That is Retail Agent Con, a half-day program on what agents are doing in retail right now. You will hear which results are real and how people and agents learn to work together for the customer and the business. The agents are here, and retail is still deciding. If you want to see what is working before your competitors do, this is the one to attend. The registration link is in the description below. This week's theme is simple: AI only works on a clean data foundation. IHL President Greg Buzek set the stage at ShopTalk Fall with a picture anyone who has driven in New York will recognize. In his words,
if the goal was to get to LaGuardia Airport from Times Square, these guys are already through the tunnel, through the toll booth, on the Long Island Expressway. Most retailers are stuck between 7th and 8th in bumper-to-bumper traffic. Getting clean, accurate picture of the truth on data has been the IHL theme for the past 8 years or longer.
Greg traces that gap back to 2018. About 15 to 20 retailers decided they were in an IT race and raised IT spend roughly 40 to 50% a year for 4 to 5 years. Today, they have the clean, centralized data that AI depends on. Everyone else is still cleaning. IHL's Intelligent Commerce research shows what that head start is worth. Winning retailers that deployed generative AI for data cleansing saw sales growth 10 times higher and profit growth 6 times higher than the rest of retail. And in several cases, the retailers with the worst data quality were pushing hardest on generative AI. Clean the data first. Snowflake made the same point this week, laying out what a reliable analytics foundation needs: accurate data, consistent metrics, and business context. Italian department store Rinascente uses Snowflake to unify customer data, refresh tables every 15 minutes, and cut report processing time about 80%. IHL's 2026 Inventory Distortion Study shows the payoff. Layering weather, sentiment, and trend signals onto clean inventory data can improve forecast accuracy by up to 50%. Retailers still fixing their foundational data cannot capture that gain, whatever their algorithm. Next, a 14-move supply chain playbook for retailers that puts RFID and real-time inventory visibility on the list next to AI. EY puts inventory accuracy at 98-99% with RFID, up from roughly 85-90% without it. IHL has measured the business impact directly. In a study of 209 brands, companies using RFID averaged 2023 profit growth 57% higher than those without it. And they expected 88% higher profit in 2024. Then there is the store itself. Orange selected Pricer to put electronic shelf labels into 561 stores in France, starting in the 4th quarter with 100 stores by year-end and finishing in 2028. Orange chose Pricer after pilots across multiple stores. IHL's 2026 Retail Transformation Study ranks electronic shelf labels first among technologies tied to sales growth, with adopters growing sales 75% higher. Leaders are also 164% more likely to already use them. And Colruyt Group in Belgium has installed 454 Toshiba self-checkout systems across 150 stores, with 51 more planned by the end of 2026. Colruyt kept its staffed lanes and skipped exit gates. IHL's study shows sales leaders are 141% more likely to be current on self-checkout. The idea that self-checkout is dead does not hold for them. Put it together. The technology exists. The leaders fund clean data, accurate inventory, and store networks that can carry robots, shelf labels, and sensors before they pick an AI tool. Greg's advice: Before you sign the next AI contract, make sure your team can tell you what is on the shelf. Now let's move to our lightning round, where we quickly recap what's happening around retail this week. Retail Insights says AI is moving retail from reporting to predictive operations, flagging inventory shortages and promotion problems before they reach the shelf. With agentic tools that act inside agreed guardrails. Whitbread, the company behind Premier Inn, replaced legacy on-premises systems with Oracle Fusion Cloud ERP and EPM across more than 950 hotels, adding self-service procurement and automatic updates. Taco Bell runs Omilia voice AI at more than 890 drive-thrus in 38 states, and Yum Brands says it helps retention. The turnover claims are self-reported, and no figures have been published. Ulta Beauty launched a ChatGPT plugin that builds personalized product bundles. Bundles of 2 to 5 eligible products get a 10% discount, and checkout happens on the Ulta site or app. Target updated its MyDay app, used by more than 350,000 team members, and says it saves more than 2 million team member hours a year. Morrison's is trialing Instacart's Caper smart trolleys in Preston, a first for a UK retailer. A survey of 1,086 Brits from Omnisend found 28% do not trust AI in retail at all. Foot Locker teamed with Stylytics on Fits, an AI closet builder in its app. Shoppers upload a photo of something they own and see outfits built around it on a virtual model. Google upgraded Universal Cart to save shopping progress. Use Google Wallet Card perks and transfer items to retailer carts at Nike, Sephora, Walmart, and Shopify merchants. Badass Coffee of Hawaii picked PAR Technology's unified suite for POS, ordering, payments, and loyalty, as it grows from about 50 US franchise locations with more than 100 more shops in development. Big Y Foods and Instacart rolled out Foodstorm kiosks and online ordering at all 79 Big Y stores. ShoeStation Group, parent of Shoe Carnival and ShoeStation, automated hiring with Fountain across more than 420 stores. Time to hire fell from 14 days to 6, and applicant drop-off fell from about 60% to 35%. Wonder is expanding Grubhub robot delivery with Avride to Jersey City, Hoboken, Arlington County, and Philadelphia. Robots now complete nearly 20% of deliveries from Wonder's Jersey City location. Booker Wholesale installed more than 500 Sony Bravia displays across 100 UK locations to launch a retail media network with Pixel Inspiration and Dunhumby. Here is the question to take with you this week. Can your team tell you right now what is on the shelf? If the answer is not a clear yes, you are among the many retailers, and that is where your AI plan starts. For all 39 stories from this week's list and free IHL white papers, head to ihlservices.com. Thanks for listening, and I hope to see you back next week.
