How Do You Build the Business Case for a Loss Prevention Program?
Build the business case on the store’s own annual loss exposure, valued at cost, rather than on industry shrink averages. For each loss type, multiply impact by frequency, then score the result with the payback, NPV, or IRR tests finance uses to judge ROI on a retail capital request. Keep apprehension counts and recovery dollars out of the loss prevention program’s proof of value.
Why the industry shrink benchmark stopped being useful
For three decades, the National Retail Federation ran an annual benchmark survey of industry-wide shrink, and its last release reported that the average shrink rate rose from 1.4% in fiscal 2021 to 1.6% in fiscal 2022, representing about $112 billion in losses in 2022, up from an estimated $94 billion the year before (lpresearch.org). NRF has not published a shrink report since. Retail Dive reported in 2024 that the organization would not release one, and NRF vice president Mary McGinty said a broad study of shrink no longer captures the industry’s key challenges (Retail Dive). What NRF publishes now is a different instrument. Its 2026 Impact of Retail Theft & Violence survey ran among senior loss prevention and security executives from February to April 2026, drew 66 retail companies representing 143 brands, and reports incident counts rather than dollar shrink: retailers experienced a 12.4% decrease in shoplifting incidents and an 8.1% decline in retail merchandise theft in 2025 compared with 2024 (NRF). NRF attributes that decline in part to retailers’ ongoing efforts to refine security measures, and adds that retailers have strengthened their ability to analyze loss patterns and assess how well different measures work against different types of theft (NRF). A falling industry trend cuts both ways for a budget request: it can make new spending look unnecessary, or make an existing program look effective when the whole market moved without it. Neither the discontinued dollar figure nor the new incident trend measures this store. The case belongs on the store’s own shrink and loss data, pulled this quarter.
What annual loss exposure means for your stores
A federal risk-analysis guideline written for computer facilities supplies the exposure formula to take into the capital meeting. FIPS PUB 65 defines annual loss exposure as the dollar impact of an event multiplied by how often it happens in a year (NIST FIPS PUB 65), and the standard notes that order-of-magnitude estimates are accurate enough for most risk analysis (NIST FIPS PUB 65), which matters since a store team rarely has exact counts, only reasonable ranges. Shrink expressed as a percent of sales moves with the sales line itself: rising sales lower the percentage even when dollar losses hold steady, according to RILA’s Total Retail Loss framework (RILA). Value each loss type at cost, the approach that framework recommends, then run impact times frequency on the actual pattern in the file, such as refund abuse at one register group. FIPS 65 also specifies that a safeguard-selection matrix should record the exposure reduction a countermeasure delivers, its annual cost, and the resulting saving (NIST FIPS PUB 65). Put each program on its own matrix row so the committee can kill one request without killing the others.
Locating the loss in your own records
Annual loss exposure only works with a real frequency estimate, and that number has to come from the store’s own records. One diagnostic comes from fraud examination practice: when cost of goods sold rises faster than sales with no change in purchase prices, quantities, or product mix, the examiner breaks the figure into beginning inventory, purchases, and ending inventory to find where the gap opened, a method that can point to theft or false inventory entries (Corporate Fraud Handbook: Prevention and Detection (p. 263)). That breakdown gives a district or store team a starting point for the impact and frequency estimates in the exposure formula. Run it by department or by store, since a single blended, chainwide number hides which locations are driving the loss. Carry the department-level gap into the exposure formula that feeds next year’s capital request.
Why apprehension counts and recovery dollars don’t prove value
Verville and Vranek offer data points pointing the same direction. At Lowe’s, Verville found that detentions averaged $184 each; Verville’s point was that 10,000 apprehensions a year still would not pay back the loss prevention salaries, compared with what the team delivered on operational, internal shrink and safety work (LP Magazine). Verville also pointed to people in the industry who reduced shrink while reducing their apprehension rate (LP Magazine). At Belk, Bob Vranek’s team made no apprehensions for the first several months of a new program, and shortage at those stores dropped sharply by the next inventory count anyway (LP Magazine), with shortage reduction over 30% in year one and about 20% in year two (LP Magazine). Recovery dollars carry a separate flaw: a recovered, resaleable item carries no financial loss once it is back on the shelf, by RILA’s own accounting (RILA). Counting that recovery as program value double-counts a loss that never happened. What Belk’s and other retailers’ asset protection leaders said about proving total retail loss buy-in covers how that argument lands with finance. Score the request on exposure reduction against a baseline. Treat apprehensions as a tactic the store may not even be allowed to use.
The three capital-budgeting tests that decide program ROI
The committee that approves a new POS controller will apply the same payback, NPV, or IRR screen to this program. Walter Palmer’s guidance for loss prevention ROI names payback period, net present value, and internal rate of return as the three most commonly used methods for evaluating a capital purchase (LP Magazine), and he recommends asking finance directly which methods, cost of capital, and approval criteria it already applies (LP Magazine). Payback period is the simplest: cost of the project divided by annual cash inflow (LP Magazine). A project costing $75,000 that returns $20,000 a year pays back in 3.75 years (LP Magazine), and Palmer notes that a payback period of three years or less is generally preferred for capital projects (LP Magazine). That result misses the three-year preference. Run the same math on your own exposure-reduction estimate before requesting the budget.
Where a program struggles to pass the cost test
Some program structures start at a disadvantage in a capital-budgeting test, however well they perform day to day. A contract store detective program sits outside a retailer’s existing payroll, so it functions as an add-on expense rather than a trade-off against an in-house position, making it economically hard to justify on cost alone (Retail Crime, Security, and Loss Prevention (entry “Contract Security in Retail Environments”)). The payback math for that kind of program will rarely clear a three-year threshold, because the cost side of the equation never shrinks: there is no existing salary line it replaces. The fix is to stop comparing it to the general shrink number and compare it instead to the exposure of the one loss type it targets, using the FIPS 65 structure above: cost against exposure reduction. If that narrower comparison still misses the threshold, scope the program smaller or drop it instead of resubmitting it with a softer number next cycle.
Why a payback number can flip in three months
A single study shows how badly timing can distort a payback claim. Researchers tracking CCTV installations found that three months after the cameras went in, the average weekly loss reduction implied a payback period of about two years (Beck and Willis). Six months after installation, the same stores’ average weekly reduction had fallen so far that the identical math implied a 58-year payback (Beck and Willis). The cameras didn’t change; the measurement window did. The ECR Retail Loss review notes there is surprisingly little independent, published evidence to help a retailer make this kind of investment decision at all (ECR). Keep the first payback figure off the request until a second window agrees with it. A comparison store is what keeps a seasonal spike from being sold as a program win.
Which metrics will make our loss prevention program look better or worse than it really is?
Apprehension counts, recovery dollars, shrink as a percent of sales, and annual loss exposure each answer a different question. Each one can send a capital request in the wrong direction.
| Metric | What it actually shows | Where it misleads |
|---|---|---|
| Apprehension count | Staff activity on the floor | Lowe’s averaged $184 per detention and still fell short of covering LP salary costs at 10,000 apprehensions a year |
| Recovery dollars | Merchandise returned to inventory | A recovered, resaleable item carries no financial loss, so counting it as savings double-counts a loss that never happened |
| Shrink percent of sales | Loss relative to revenue | Rising sales lower the percentage even if dollar loss holds steady |
| Annual loss exposure | Impact dollars times frequency for one loss type | Still an estimate, though an order-of-magnitude estimate is accurate enough for most risk analysis |
Report exposure reduction as the headline number for the committee, and keep the other three as supporting detail rather than substitutes for it.
Put it together: the one-page business case
Arrive with a one-page sheet that places exposure reduction next to program cost in finance’s usual capital-request layout. Frame those figures in total retail loss, not as a shrink percentage.
- Pull the store’s own loss and shrink data for the past four quarters instead of a published benchmark, run a store-level loss survey, and list specific loss types with rough frequency
- Convert any retail-valued losses to cost using the store’s cost-to-retail ratio
- Calculate annual loss exposure, impact times frequency, separately for each loss type
- Ask finance which capital-budgeting method, cost of capital, and approval threshold it uses
- Set a measurement plan with a comparison store and at least two reporting windows, locked in before the pilot launches, not after the first results come in
Frequently asked questions
What's the difference between shrink and annual loss exposure?
Shrink lacks one agreed industry definition, which is why cross-company shrink surveys are hard to benchmark against each other, according to the Retail Industry Leaders Association's Total Retail Loss framework. Annual loss exposure sidesteps that problem by scoring a single, specific loss type in dollars.
Should a loss prevention budget request include projected apprehension numbers?
No. In 41% of retailers surveyed, no employee is authorized to stop or apprehend a shoplifter (lpresearch.org), so a program built around arrests may not even be available to many of the stores it protects. Score it in averted loss.
How long should a pilot program run before reporting results?
There's no fixed rule, but one early measurement is not enough. One CCTV study found a payback estimate near two years after three months, then the same system implied a 58-year payback just three months later (Beck and Willis). Measure across more than one inventory cycle before calling the result.
If finance uses net present value or internal rate of return instead of payback period, does the math change?
No. The exposure reduction a program is expected to deliver becomes the input either way: the cash inflow in a payback calculation, or the yearly return discounted against the cost of capital in an NPV or IRR model. Ask finance which method it applies.
Do falling theft incident counts mean a loss prevention program is working?
Not on their own. NRF's Impact of Retail Theft and Violence 2026 report found that respondents saw fewer shoplifting incidents on average between 2024 and 2025, and NRF says security efforts explain only part of the drop. An industry-wide drop helps stores with and without a program, so judge yours against a comparison store.