In May 2026, a major retail technology startup reached a $1 billion valuation after raising $170 million to scale its RFID-based inventory platform. The milestone confirmed what most retailers already know: inventory accuracy is a real, expensive, unsolved problem.
But the funding round also surfaced a question worth asking carefully. Knowing where every item is located inside a store is not the same as knowing why that store underperforms. And the two problems require different tools.
The problem RFID solves
RFID technology attaches a small radio tag to each item in a store's inventory. Ceiling-mounted readers scan those tags continuously, giving store managers a real-time view of exactly where every tagged product is located. When a customer asks for a different size, a staff member can see on their device whether it's on the floor, in the fitting room, or in the back.
This solves a problem that has plagued retail for decades. Buy-online-pickup-in-store (BOPIS) orders get canceled because a system shows an item as available when it isn't. Distribution centers ship 80 units but a store accepts 100 without checking. Inventory counts happen quarterly rather than continuously. RFID addresses all of this directly.
The results are real. Retailers using RFID report significant reductions in BOPIS cancellation rates, faster replenishment cycles, and improved inventory accuracy. These are legitimate operational gains.
What RFID does not do is tell you anything about the store itself as an operating environment.
What RFID cannot see
An RFID system knows that 47 units of a size-medium shirt are in the store right now, distributed across the floor and the stockroom. It does not know:
- Whether a customer waited 12 minutes for assistance and left without buying
- Whether the peak traffic hour between 1pm and 2pm coincides with the lowest staff coverage of the day
- Whether the brand's visual standards are being maintained or the fitting room area has deteriorated since the last manager walk
- Whether the store's NPS score dropped because of a service experience problem, not a product availability problem
- Whether the busiest zone of the store has no staff presence on Thursday afternoons
These gaps matter because most lost revenue in retail is not lost to stockouts. It is lost to friction. A customer who can't find help. A fitting room that signals neglect. A checkout queue that pushes someone toward the exit. These events don't appear in inventory data. They don't trigger RFID alerts. They accumulate invisibly until they show up as declining conversion and a falling NPS score that nobody can explain.
Inventory accuracy tells you the product was there. It cannot tell you whether the customer experienced a store that made them want to buy it.
What store intelligence actually means
Store intelligence is operational analytics built around how a store performs as a human system. It answers questions about people, behavior, and execution rather than product location.
The data source is different from RFID: computer vision platforms use the cameras already installed in most stores, analyzing video feeds in real time without storing footage or collecting biometric data. The questions they answer are operationally distinct:
- Traffic and flow: How many people entered the store, when, and where did they go? Where are the conversion drop-off points?
- Service time: How long are customers waiting in queue? Is the fitting room understaffed during peak hours?
- Staff coverage: Are associates present in the zones where customers need help? Are opening and closing procedures being followed?
- Brand execution: Is visual merchandising being maintained? Are marketing materials displayed correctly across locations?
- Behavioral shrink signals: Are there patterns at the point of sale or in high-value zones that correlate with inventory loss?
The platform doesn't track tagged items. It tracks what the store is doing with the people who walk into it.
Two signals, two decisions
The clearest way to separate these two categories is to look at the decisions each one informs.
Answers: "Where is the product?"
- Is the size a customer asked for on the floor or in the stockroom?
- Did the shipment arrive complete or short?
- Is inventory accurate enough to fulfill BOPIS orders?
- Where in the supply chain is the shrink occurring?
Answers: "How is the store performing?"
- Why are customers leaving without converting despite available inventory?
- Which stores underperform similar-inventory locations?
- Are service standards consistent across the estate?
- Is the customer experience matching the brand promise on the floor?
A retailer can have perfect RFID coverage and still lose significant revenue to service failures, layout problems, and execution inconsistency. These are not the same problem. They are not solved by the same tool.
Side-by-side comparison
| Dimension | RFID inventory tracking | Store intelligence (computer vision) |
|---|---|---|
| Core question answered | Where is the product? | How is the store performing? |
| What it tracks | Tagged items | People, behavior, operations |
| Hardware required | RFID tags on all inventory, ceiling readers | Existing cameras (no new infrastructure) |
| Implementation complexity | High: tagging all inventory, reader installation, POS integration | Low: connects to existing camera feeds |
| Primary beneficiary | Operations, supply chain, store associates | Store operations, CX, loss prevention, field teams |
| CX insights | Indirect (inventory availability affects experience) | Direct (queue times, service rates, traffic patterns) |
| Shrink coverage | Inventory discrepancies, supply chain shrink | Behavioral shrink signals, POS anomalies |
| Biometric data risk | None (tracks items, not people) | None with behavior-only platforms (no facial recognition required) |
Which problem to solve first
The right starting point depends on where revenue is being lost.
If your primary pain is inventory inaccuracy: BOPIS cancellations above 10%, recurring stockouts in fast-moving sizes, or supply chain discrepancies you can't trace, RFID addresses those problems directly. The investment is significant, but the ROI case is straightforward.
If your primary pain is unexplained performance gaps between locations, declining NPS scores with no clear operational cause, or inconsistent service quality across your estate, store intelligence addresses those questions. It doesn't require tagging inventory. It connects to the cameras you already have.
Many retailers find they need both over time. They are not competing investments. But they should be evaluated against separate problems, not compared as alternatives to the same question.
The retailers gaining competitive advantage from in-store technology right now are not choosing between inventory accuracy and operational intelligence. They are sequencing the investments correctly, starting with the problem that is actually costing them the most revenue today.
Before committing to either category, map the actual revenue loss. Count the customers who left without converting during a peak service window. Measure the variance between your highest and lowest-performing stores. The answer to "which problem first" is almost always visible in that data, if you have it.
If you don't have it yet, that's the first thing store intelligence gives you.
EdgeRetail is a store intelligence platform built on computer vision. It connects to your existing cameras to surface traffic, service, brand execution, and shrink signals across your locations. No new hardware. SOC 2 certified and GDPR compliant. Learn more about EdgeRetail →