What five conversations across the Retail Razor Podcast Network reveal about AI in retail and how AI should serve people
For two years, the loudest story in retail AI has been the same one. The machines are coming for the jobs, so cut headcount fast and call it efficiency. Across five episodes this season, we kept running into the opposite. Different guests, different shows, different rooms, and the same answer every time. AI should serve people - not replace them.
That's not a slogan. It's the pattern that showed up in the data. It showed up in the leadership stories. And it came through on stage at two very different conference floors. Here's the evidence, episode by episode.
The retail AI data says the opposite of the headlines.
We'll start with the numbers that should settle this.
On Data Blades, Casey and I went back through IHL Group's 2026 Retail Transformation Study with Greg Buzek. If AI in retail were quietly deleting jobs, the most AI-ready retailers would be shrinking their store teams. The data shows the reverse. They're growing store headcount at 70% versus 36% for their peers. The fastest-growing retailers are 115% more likely to be hiring store associates rather than cutting them.
Greg calls the model the augmented associate and it shows how AI should serve people rather than replace them.
The retail winners use AI as a tool to make store associates more capable, then keep hiring those associates because they're now worth more. He traced the pattern back to self-checkout 25 years ago.
Kroger trained its best people to teach shoppers and moved them to high-margin fresh counters.
Kmart bolted in the machines and cut the labor.
Same technology, opposite philosophy, opposite outcome. The augmented associate is simply that lesson, updated for the AI era.
Why AI projects fail starts at the top, not the frontline.
The data says keep your people.
One of the most experienced voices we spoke to this season explained why the alternative keeps failing. On Retail Transformers, we interviewed Julie Averill, the former global CIO of Lululemon who helped scale the company from $2 billion to $10 billion and author of Chief Impact Officer.
Her framing reshapes the entire retail AI transformation conversation:
AI doesn't fix your culture, it reveals it.
When a retailer layers AI on top of broken processes, weak data, and unclear decision rights, the result is a flawless demo. But useless output. Like the AI that recommended stocking winter coats in Miami.
That, in Julie's view, is why AI projects fail at a rate as high as 95 percent, per a recent MIT study. The technology is rarely the point of failure.
For retail leaders, the lesson lands directly on people. Treating retail AI as a quiet route to cut headcount mistakes the tool for the strategy. It's a fast path to the pile of reasons why AI projects fail. Real retail AI transformation is human work. Honest leadership about what's changing. The unglamorous data foundation that makes everything else possible. And developing teams so capable that leaders make themselves unnecessary. It's a powerful reason why AI should serve people and not replace them.
Understanding why AI projects fail is the first step toward retail AI transformation that actually serves people. As Julie puts it, AI is not the point. The people are.
Future-proof retail skills get built on the store sales floor.
If the data and a veteran CIO both say AI should serve people and not replace them, the next question is clear. What should those people be doing? On Blade to Greatness, Ron Thurston, author of Retail Pride and Human Pride, gave the clearest answer we've heard. The most future-proof skills in retail aren't technical.
They're human.
Communication, empathy, adaptability, and the ability to earn a stranger's trust in 90 seconds.
Frontline workers build those skills on the sales floor. One conversation at a time. Not in a corporate binder.
Ron's view of AI in retail is the constructive one. Let the technology carry the operational weight, the policy look-ups and product specs, so frontline workers can focus their attention on the customer in front of them. The technology buys back time. How a retailer spends that time decides everything. And that's how AI should serve people in stores instead of replacing them.
The Lead Summit drew the same line in the sand.
We saw the same split in person. At The Lead Summit in New York, every brand in the building was investing in AI, so that part of the conversation was over. The question that separated the strong sessions from the noise was simpler.
Where does the human stay in the loop?
The brands pulling ahead used AI to buy back the hours their teams lose to busywork, then aimed those hours at the parts of the experience a machine can't fake. The brands automating the human connection out of the experience will spend next year wondering where the loyalty went. And that means, you guessed it, once again, AI should serve people and not replace them!
Zebra Technologies ZONE 2026 proves the tool makers agree.
The clearest signal came from the company that builds the tools. At Zebra Technologies ZONE 2026 conference in Nashville, I sat down with two executives expecting a hardware pitch and found a people story instead.
Two stats framed the whole event:
frontline workers spend roughly 30 percent of their time hunting for information they can't find, and about half of their errors trace back to not understanding a standard operating procedure.
A better scanner fixes neither. Retail AI that answers the question in the flow of work hands a third of the day back.
CTO Tom Bianculli called it pragmatic AI for the frontline, with a nursing example that sticks. Clinicians lose a quarter of every shift to documentation. Ambient AI on the device gives much of that time back.
James Poulton, who leads Zebra's mobile computing business, put it plainly. It's a fallacy to believe frontline workers don't want tools. Zebra's own research shows that giving store associates good technology makes them happier and more likely to stay.
The augmented associate again, straight from the company arming the frontline. It's yet another example of why AI should serve people and not replace them!
Five rooms, one conclusion for retail AI.
These weren't five people coordinating a message. Greg is reading research. Julie led tech for a $10-billion-dollar retail organization. Ron is coaching retail leaders. And the two events sat a world apart. They all reached the same conclusion about AI in retail. AI should serve people - not replace them.
The retailers winning right now?
They treat AI as a way to make store associates more capable, more available, and more valuable.
The ones struggling treat it as a reason to employ fewer frontline workers. The technology is identical. The philosophy is the whole game. As Casey put it on the ZONE recap episode, your associate experience is your customer experience. The technology is finally catching up to that truth.
So if you lead a retail team weighing an AI budget against a headcount line, here's the question worth sitting with.
Are you buying a tool to serve your people, or an excuse to replace them?
The data already told you which one pays off.
AI should serve people - and the retailers who understand that are the ones still standing in five years.
Hear the full conversations on the Retail Razor Podcast Network: The Retail Razor Show, Retail Transformers, Blade to Greatness, and Data Blades. Subscribe wherever you get your podcasts, and join the conversation with Ricardo Belmar and Casey Golden.
