Retail finally found a theft control that works. The locked case does exactly what it promises: the product stays on the shelf until someone with a key decides otherwise.
The problem is what else it stops. 60% of American shoppers now encounter locked-up merchandise on a regular basis, and 61% say the amount grew in the past year (Numerator, 2024). The industry deployed its most aggressive in-store control in decades — and never published a number for what it costs in sales.
The number exists. It is just not in LP's reporting. It is in the patent filings of the companies selling the fix.
This edition runs the math the lock never had to pass.
The control the customer pays for
The locked case is the only loss prevention control the legitimate customer experiences personally. The data on how they respond to it:
27% of shoppers will not wait for the key — 17% switch retailers and 10% abandon the purchase entirely when they hit a locked product (Numerator, September 2024, 5,000+ consumers surveyed).
Fewer than one in three shoppers even summon an employee — and among those who leave without buying, most report buying elsewhere rather than picking an alternative in your store (Consumer World, 2024).
The friction is concentrated where margins are thinnest — drug and mass retail, the channels where lockups grew fastest, are the same channels bleeding trips to online (Numerator, 2024).
Every one of those events is invisible in shrink reporting. The case protected the item. The customer just never came back.
The vendors wrote the number down
As with self-checkout (Edition #10), the honest math lives in patent filings — where a vendor must describe the problem accurately in order to sell the solution. The US patents behind self-service locked-case technology state it plainly:
The process of accessing locked merchandise "typically results in a loss of 25% to 50% of sales" in the protected category, with the value of lost sales often exceeding the theft prevented (USPTO filings).
Retailers "inconvenience 99% of legitimate shoppers to stop theft from the 1%" — the vendor's framing, not mine (USPTO filings).
Run the trade on a category doing USD 1M with 3% shrink: the lock saves at most USD 30K of loss. At the patent's own 25% sales-loss floor, it removes USD 250K of revenue — roughly USD 87K of margin at 35% gross. You paid nearly three dollars to save one.
Joel Bines, the retail advisor and former AlixPartners managing director, calls locked cases "the single most customer unfriendly" strategy in retail. The industry's answer so far has been to make the key digital.
Why nobody stopped it
The lockup wave survived five years without a business case for the same reason self-checkout did: the two sides of the trade land in different rooms.
The shrink saved lands in LP's line. The sale lost lands in merchandising's — as soft demand, unexplained velocity decline, a category problem. No major retailer has disclosed sales lost to security measures.
The fixes concede the point. Walmart is piloting NFC unlock through its own app in a few hundred stores; CVS is piloting unlock for ExtraCare loyalty members in New York (Modern Retail, Bloomberg). When you need to build an app so customers can reach the product, the control was mispriced.
Trust is becoming a loyalty tier. App-gated unlock means your best-known customers get access and everyone else gets the call button. That is a segmentation decision with brand consequences nobody modeled.
A perspective from the field
I have signed off on keeper boxes, cages, and locked cases in four countries. Not once did the approval include a sales impact estimate — only the shrink we expected to save. We measured the win in the line we owned and never looked at the line we didn't.
What changed my view was standing in an aisle watching a customer press the call button on a case. He waited. He checked his watch. He left. The case had a perfect record that day: nothing was stolen. Nothing was sold, either. A control that defeats the thief and the customer at the same rate is not protection. It is a closed sign with extra steps.
Three practical moves for the next 90 days
Run the two-line P&L on every locked category this quarter. Sales velocity before and after lockup versus matched control stores, times gross margin, against the shrink actually avoided. Where margin lost exceeds shrink saved, the case is a self-inflicted loss. That store-and-category list is the deliverable.
Put a service SLA on the key. If the case stays, staff it: measure time-to-unlock and response rate the way you measure register queues. A one-in-three summon rate is not customer laziness. It is a service failure with a shrink alibi.
Protect SKUs, not categories. ORC targets are specific — high value, high resale, concealable. Lock those. Tier everything else down to wraps, shelf alarms, or smart cases, and pilot any unlock technology with matched-store measurement before chain-wide rollout. Edition #10's lesson applies to the cure, too.
Closing note
The locked case is not always wrong. For a USD 400 power tool with an active resale market, the math clears easily. The failure is treating the lock as free — as if the only line it touched were shrink.
Loss prevention spent two decades earning a seat at the merchant's table by talking about sales, not just loss. The lockup wave is where the industry quietly handed that argument back. Take it again: run the number, publish it internally, and let the P&L — not the fear — decide which cases stay.
I want to hear from leaders who have measured sales transfer on locked categories. Did the 25–50% number hold in your stores? Reply with anything you can share, anonymized if you prefer.
Forward this to one LP or AP leader who should be reading it.
— Gabriel
The LP Brief is a weekly intelligence read for senior loss prevention and asset protection leaders. Free. No vendor noise.
Not yet subscribed? thelpbrief.com