A Senior Retail Exec's View on Store Intelligence

A senior retail executive told me recently he didn't want a slide deck. He wanted to walk the store. "Let's meet in the store," he said, “and I’ll show you the challenges I’m facing”.

So we did.

We went aisle by aisle. He showed me what tasks a store associate sees on their device when they first arrive. He talked to me about how customers react when something's wrong with a shelf. He highlighted the job a picker was doing, trying to find products that were in secondary locations, and how long it was taking them.

It’s one of the best customer conversations I’ve ever had. Not only for having an in-depth discussion. But because I could see what was happening in real time.

It was retail in miniature.

Depending on the format, 60-90% of a retailer's revenue and profits are generated within its stores. And yet the store remains, in my view, the biggest blind spot of the business in retail. Not the supply chain. Not the e-commerce site. The store floor itself.

And that blind spot has a price. Global inventory distortion, the combined cost of out-of-stocks and overstocks, as reported by IHL, runs to $1.73 trillion a year. About 5% of global retail revenue is sitting there unclaimed.

The day-to-day reality behind that number isn't pretty either.

From our work with retailers, we see that on-shelf availability sits below 90% almost everywhere. Price labels are wrong roughly 10% of the time. And shelves rarely match planograms.

One retailer we work with ran a pilot and found planogram compliance sitting at around 60% across their stores. They were shocked enough to stop the meeting to raise the issue internally.

It's not a supply chain problem

This is the part that catches most retail leaders off guard, and it caught me off guard, too, the first time I saw the data. When you look closely at on-shelf availability gaps, 50% to 70% of that "missing" stock isn't missing at all.

It's in the back room. It's in top stock. It's sitting in the receiving bay because it was delivered an hour ago, and nobody's moved it yet. Some of it is shrinkage, stolen, damaged, or lost to a bad count. But most of it is simply misplaced, not missing. Nobody in the store can see where it actually is or tell the two apart.

The store is constantly watched, but almost none of that watching translates into understanding. Watching isn't knowing, and knowing isn't the same as doing something about it.

Closing that gap, turning what gets watched into what gets understood and acted on, is where we introduce the concept of store intelligence.

Vision AI: The enabler of store intelligence

Full store intelligence isn’t possible without cameras and AI.

We call the combination vision AI. It's machine learning, augmented reality, and computer vision fused into a perfect view of the shelf, the price label, and the product, at any point during the day.

Our customers capture that view in a few different ways:

• An associate walking and recording with a smart device.
• Fixed shelf-edge cameras watching high-traffic sections.
• Robots that roam large-format stores on their own.

None of these is the "right" way to do it. They're just different ways of monitoring shelves that adapt well to different store formats, sizes, and data capture frequency needs.

And even this part of the process can save time and help with labor efficiency.

A video-based capture process runs about 3 times faster than a manual gap scan, and finds roughly twice as many gaps as a well-trained human scanner.

What used to take 2 to 3 hours to walk an entire store now takes 40 minutes. Leaving more time to focus on fixing the issues that are found.

Store intelligence: turning sight into action

Capturing the shelf is the easy half. The harder half, and the part I spend most of my time thinking about, is turning shelf-level intelligence into the right task, for the right person, at the right moment.

Not a report that lands in an inbox every week.

An alert that tells an associate to restock modules 4 and 5 in aisle 7 right now, or sends someone into the pasta aisle because the planogram has drifted, or flags a cheese that needs marking down before it expires.

And when several of those alerts land at once, which happens on most shifts, the platform decides which to tackle first. A missing high-margin SKU jumps ahead of a label mismatch in a quiet aisle. The associate isn't left guessing which fire to put out first.

That kind of call is only possible because our store intelligence platform isn't a collection of separate tools bolted together. It's built as modular solutions that stand alone but are designed to run as one.

What is a store intelligence platform?

A store intelligence platform brings together shelf and inventory data, on-shelf availability, price, and planogram compliance, expiry dates, and loss prevention signals, into one shared data layer instead of siloed systems. It turns captured data into prioritized, actionable guidance so store teams focus on what matters most and execute, with results compounding across shrink reduction, labor efficiency, and sales.

Consider:

• Gap scanning and replenishment: An associate walks the aisle once, and vision AI flags missing items, pushing on-shelf availability to over 95%.
• Price and planogram compliance: AR-guided checks show an associate exactly where and how to fix a problem and log the fix as it happens, increasing compliance by 20%.
• Expiry management: The same capture process reads dates in 60% less time than typing them in by hand and triggers markdowns automatically, cutting waste by up to 40%.
• In-store picking: Knowing exactly where every item sits turns a pick list into a routing decision, growing fulfillment capacity by about 5% without the need for more labor.

Run separately, each one pays for itself.

Run together, off one shared picture of the store, they compound.

Previously separate, time-consuming manual processes are now eliminated. Resulting in big labor gains. For example, capturing on-shelf availability data allows the creation of a digital twin of your entire store. From there, you can assess price and planogram compliance, generate picking routes, and provide guidance to further improve labor efficiency.

One shared intelligence layer, 4 jobs.

And it doesn’t stop there. With advances in vision AI, the most common theft methods can be detected and loss prevented.

Captured product facings during shelf monitoring train self-checkout systems to spot item-switching.

And in-aisle cameras can spot shelf sweeping, where several items leave the shelf at once. If nothing gets paid for, the system alerts staff or corrects the stock count on its own, clearing phantom stock before it spreads.

If you connect what shelf cameras see to what the POS actually rings up, the loop closes. What comes in and what goes out is reconciled.

Why does this beat the big transformation project?

The same executive who walked me through his stores made one more point that remains top of mind.

He wasn't interested in a big bang. Sweeping change looks great from a stage or in a video. It rarely survives contact with thousands of stores and thousands of employees. His own words: "I prefer approaches where I can tweak and adapt my current processes."

That's the model I'd point any retailer toward. And it’s exactly what we built the platform to do.

Take the gap scanning and cycle counting your stores already run, and add a vision AI layer instead of ripping it out and starting over.

Most retailers we work with can build a real business case, with real numbers, off a four-week pilot and nothing more than a CSV export of their product catalog. No system integration is needed to get started.

The store isn't short on things happening. Inventory arrives at it, associates walk it, and shoppers buy from it.

What's missing is the bit in between: seeing what's happening on the store floor clearly enough to fix it before a shopper does.

That's what store intelligence is all about, and it's about as low-risk an AI investment as retail gets right now.

Real-time inventory visibility and guided actions

Scandit store intelligence

Explore