Here is a sentence that should make every AP leader uncomfortable.
A store's apprehension count can rise, quarter over quarter, while that same store's shrink also rises.
The two numbers are not locked together the way the AP scorecard assumes. One measures how busy the loss prevention team is. The other measures how much the store is actually losing. AP reports the first one up the chain. The CFO only cares about the second.
That gap — between the metric AP optimizes and the metric the business pays for — is the quiet structural problem at the center of the function.
The numbers AP reports up the chain
Walk into any AP operating review and you will see the same scorecard: apprehensions, recoveries, cases closed, cases referred. These are the metrics of the function. They are also, almost entirely, metrics of activity, not outcome.
Look at what they actually add up to:
U.S. retailers apprehended an estimated 350,000 shoplifters and dishonest employees in 2022 — a 45.6% increase over the prior year (Jack L. Hayes International). Shrink did not fall 45.6% in response. It rose.
The industry-wide recovery rate on theft losses is roughly 10.9% (Jack L. Hayes International). For every dollar walked out the door, about eleven cents comes back.
For internal theft specifically, the average recovery is $0.18 on the dollar. The case file, the interview, the investigation hours — all of it, to recover under a fifth of the loss.
A rising apprehension count tells you the team is working. It does not tell you the store is winning. Those are different questions, and the AP scorecard is built to answer the first one.
The deterrence assumption that isn't tested
The theory underneath the apprehension metric is deterrence: catch offenders, and future offenders are discouraged. It is a reasonable theory. It is also rarely measured.
What the data shows about the enforcement chain:
64% of retailers report less than half of store theft to law enforcement (NRF 2025), citing lack of police response as the primary reason
Roughly 60% of retail crime goes unreported because retailers have lost faith in prosecution
About 40% of U.S. retailers authorize no employee to apprehend a shoplifter at all — the entire model has shifted to evidence collection and off-site handling
So the deterrence chain that justifies the apprehension metric is broken at nearly every link: most theft is never reported, most reports never reach prosecution, and a growing share of stores have removed apprehension from the floor entirely. The apprehensions that do happen are a thin slice of total theft — and there is little evidence that slice meaningfully deters the rest.
The metric persists not because it has been proven to reduce loss, but because it is easy to count. Apprehensions are discrete, visible, and reportable. Deterrence is diffuse, invisible, and hard to measure. The function optimizes what it can count.
The cost the scorecard hides
Here is the part that should reach the C-suite.
The apprehension scorecard not only fails to capture deterrence — it actively hides the cost of the tactics used to generate the numbers. The defensive measures that protect against the theft AP is counting are themselves removing sales from the floor:
30% of retailers closed store locations in response to retail theft (Appriss 2026)
65% of retailers removed specific products from the sales floor entirely to avoid theft — those are sales removed, not just shrink prevented
27.6% of shoppers say they would stop buying from a retailer where they felt unsafe (Appriss 2026)
A locked case, a removed SKU, a heavy uniformed presence, a closed store — each of these reduces the shrink number AP reports. Each also reduces revenue, in ways that never appear on the AP scorecard. The function gets credit for the shrink it prevents and pays no penalty for the sales it costs. The two numbers live in different reports, owned by different people, and almost never get netted against each other.
When a CFO eventually does net them, the AP leader who has only ever reported apprehensions and recoveries has nothing to say.
A perspective from the field
Early in my career I ran a security operation that was proud of its apprehension numbers. They went up every quarter. We celebrated them. Leadership celebrated them.
It took me too long to notice that the stores with the highest apprehension counts were not the stores with the lowest losses. They were often the stores with the most theft — which is exactly why there was so much to apprehend. We were measuring the disease and calling it the cure.
The shift that changed everything was moving from counting what we caught to measuring what changed: shrink as a percentage of sales, trend over time, netted against the revenue cost of the controls we deployed. The day I walked into a review with that number instead of an apprehension count was the day the CFO started treating security as a business function instead of a cost center.
Three practical moves for the next 90 days
Separate your activity metrics from your outcome metrics — on paper, explicitly. Apprehensions, recoveries, and cases are activity. Shrink as a percentage of sales, trend over time, is outcome. Report both, clearly labeled, and never again let an activity number stand in for an outcome the business actually cares about.
Net your prevention against your revenue cost. For every major control — locked cases, removed SKUs, store closures, guard presence — estimate the sales impact alongside the shrink reduction. You will not get it perfectly precise. You do not need to. A rough, honest netting is infinitely more credible to a CFO than a shrink number reported in isolation.
Bring one deterrence question to your next review that you cannot currently answer. Does our apprehension activity actually reduce subsequent theft in the same store? If you do not know — and almost no one does — that is the research project that turns AP from a function that counts events into a function that produces evidence. The leader who runs that study owns the most valuable number in the building.
Closing note
Apprehensions are real work. The investigations are real, the recoveries are real, the offenders caught are real. None of that is theater.
The theater is the reporting — the quiet substitution of a number that is easy to count for the number that actually measures whether the store is winning. AP has spent decades getting very good at producing the first number. The next decade belongs to the leaders who learn to produce the second.
I'd like to hear from AP leaders who have moved their reporting from activity to outcome. What changed in how leadership treated you? 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