Here’s why boring wins as RTIH launches its first ever Retail Technology Hot 100 List this week
Ahead of the first ever RTIH Retail Technology Hot 100 List going live tomorrow, Holland & Barrett’s Group CFO, Vineta Bajaj, a member of the judging panel, argues the most bullish signal in this year's list is that the hype has quietly moved from the front of the shop to the back.
I have judged enough of these lists to know when the ground shifts. Two years ago almost every entry for award shortlisting was a chatbot with a nice interface and a demo that fell over the moment you asked it a second question.
This year the companies that made me sit up are the boring ones. The ones fixing whether the product is on the shelf, whether the stock in the van is where the system says it is, whether the robot in the warehouse pays for itself. So why is boring a compliment?
The tech entries that earned their place this year have one thing in common. They run in the background. They handle the routine 80% of a task and surface only the exceptions a human should think about, and they do it without a dashboard or a launch event. That is the opposite of the copilot-and-chatbot wave that has dominated many shortlists, most of which has since gone quiet.
The reason is simple and slightly brutal. Very few of those companies were ever clear on the problem they were solving, and fewer still had the data discipline to solve it. You can bolt the cleverest model in the world onto a broken process and a messy data set, and all you get is a faster version of wrong. Shit in, shit out. The founders who understood that are the ones still standing.
Vineta Bajaj: One should judge technology by the size of the problem it attacks and whether it can prove that, in the real world, it pays.
The second shift is that autonomy stopped being a science project. For years the robots, the drones and the computer vision stores on these lists were sold on the wow. You watched the video, you nodded, and then you asked what it cost to run and the room went quiet. That changed this year.
The strongest entries in warehouse robotics, shelf scanning and last mile delivery now lead with the number that matters: cost per pick, cost per drop, hours of manual counting removed.
One of them can tell you it delivers several pounds cheaper than a human courier, which is the difference between last mile being a permanent drain on margin and an actual advantage. I spent a decade around warehouse automation, so I know how rare it is to see this category talk about economics before it talks about hardware. When it does, pay attention, because that is the sound of a technology crossing from interesting to inevitable.
The third theme is the one that most founders get wrong when they pitch me. Everyone now has access to the same frontier models. That means the model is not the moat, and a thin wrapper around someone else's model is not a business. The defensible companies on this shortlist are doing something harder and far less glamorous. They are taking an old, unloved category, workforce management, replenishment, store execution, and rebuilding it from the ground up with AI in the foundations rather than bolted on the side.
Better still, they own a flywheel of real production data that improves with every shift, every pick, every store. That data is the moat, and you cannot buy it or prompt your way to it. It has to be earned in the messy real world, which is exactly why it is worth something.
But of course, everyone will ask about the thing that is starting to sound the most interesting, especially for retailers. Agentic commerce, the AI that discovers and buys on the shopper's behalf is coming or rather, is here! And it will reshape the funnel when it arrives. I judge on production, not on promise, so I kept my enthusiasm in check.
The companies claiming agentic capability today are mostly showing it in a controlled sandbox, not running it across a live estate at peak on a wet Tuesday in November. That’s a reminder to tell the difference between a pilot and a business. Ask me again next year and I suspect this becomes the headline, but I don’t think it is there yet en masse.
One more thing the shortlist put to bed. The store is not dying, its role and job is changing. The most interesting store technology this year was not trying to replace the shop with a screen. It was making the physical estate work harder, cheaper and smarter, because the people building it understand that a store is still the highest converting, most profitable channel most retailers own.
So here is what this year taught me, and it is the same lesson I apply running multiple global finance functions. One should judge technology by the size of the problem it attacks and whether it can prove that, in the real world, it pays.
The flashy front-end that dazzles in a boardroom is rarely the thing that moves your cost base or your cash. The boring company fixing an expensive, structural problem it can actually measure almost always is. Think like an owner rather than a spectator, and you stop being seduced by the demo and start being interested in the deployment.
The most exciting retail technology I saw this year does not look exciting at all. That is precisely why it will win.
