To what extent do you believe AI is delivering real, measurable value in retail today, versus being driven by expectation or hype?
AI delivers real value but only when built on real data. In conversations with retail leaders, I see a growing divide between vendors making AI claims and vendors proving them. Buyers are getting sharper at telling the difference. What cuts through isn’t a count of AI models. It’s specificity about what the AI does, transparency about how it works, and outcomes tied to real business problems. For Appriss, that’s an auditable Data Intelligence foundation based on 40% of U.S. retail transactions, 20+ years of transaction data, and return decisions made in under a second at 99.99% accuracy across in-store and online.
Where do you see the greatest opportunity for AI to transform retail — customer experience, operations, supply chain, or decision making?
It’s decision making. But the value isn’t the same across every category, and that matters.
In returns, the opportunity is accuracy. Our 2026 Total Retail Loss Benchmark Report put total return volume at $706 billion, with $100 billion of that lost to preventable fraud and abuse. Getting that call right — approving a loyal customer without friction, declining a fraudulent one without a rule being written — requires cross-retailer behavioral patterns across 250 million consumer identifiers that no single retailer’s dataset can see. That’s what drives 99.99% accuracy on returns decisions. Speed is easy. Accuracy at that scale is the challenge.
In shrink and operational loss, the opportunity shifts. The same report identified $66 billion in preventable employee theft, inventory errors, operational errors, and ORC. Surfacing the right anomalies faster, such as a cashier’s void rate four times their peer group, incidents connecting across locations, cuts investigation time from 45 minutes to 5. Both are AI. Different problems, different values. The retailers who understand that distinction are getting the most out of technology.
In your view, what is currently the biggest barrier to successful AI adoption in retail — technology, data, skills, culture, or governance?
Honestly, it’s a tie between data and culture. The data challenge is structural because no single retailer’s dataset, however large, can see the cross-retailer behavioral patterns that consortium intelligence reveals. But the cultural challenge is just as real. I talk to LP leaders regularly who are still running their operations the way they were 20 years ago, and introducing AI requires them to think fundamentally differently about their role. Technology is ready; getting organizations to trust it and act on it is the harder problem.
How important is explainability and transparency to you when evaluating AI-driven decisions in retail?
It’s the foundation of everything, and it’s a commercial requirement, not just an ethical one. The LP leader is rarely the only voice in a sales cycle. IT, finance, and internal audit all have questions about how decisions are made, where data is stored, and how models are retrained. If you can’t answer those questions with confidence, you don’t get the deal.
For us, transparency is operational. Every return authorization decision we make is fully auditable — the factors, the model version, the policy applied. All consequential AI outputs support human review workflows, and retailers retain full control of final decision authority. These policies are built into the platform. That’s what earns the trust to let AI operate at the point of transaction.
Do you believe current regulatory developments will accelerate responsible AI adoption or slow innovation?
Smart regulation raises the floor, and that’s good for platforms already built above it. We have a formal AI governance program, an AIGP-certified General Counsel, and fully auditable records for every return decision we make. Clearer standards make those conversations easier, and enterprise buyers are already asking these exact questions before they sign.
