What Is Retail Shrink? Definition, Formula, and Causes
What Is Retail Shrink?
Retail shrink is the difference between what a store’s inventory records say it should have on hand and what a physical count actually finds, expressed as a percentage of sales for the period measured. NRF’s National Retail Security Survey put the average shrink rate at 1.6% of sales in FY2022, representing $112.1 billion in losses (NRF). In 2024, NRF said that survey still held the latest available shrink statistics (Retail Dive).
Shrink vs. Total Retail Loss
Shrink only counts what a physical inventory count can see: the gap between the ledger and the shelf. Total Retail Loss counts loss the physical inventory gap does not, across the sales floor, the supply path, and headquarters. ECR’s Total Retail Loss report proposes a new definition built on a 33-category typology spanning the entire retail environment, from shop theft on the sales floor to fraud at corporate headquarters (ECR). The typology organizes loss around four centers: stores, retail supply chain, e-commerce and corporate. It also splits loss first into known and unknown, then splits known loss again into malicious and non-malicious categories (ECR). Malicious loss is intentional diversion of goods, cash, services or profit. Non-malicious loss is cost without that intent (ECR).
One thing the typology deliberately leaves out: markdowns, price matching and customer compensation. ECR’s report treats those as the ordinary cost of running a retail business, not as losses, and recommends keeping them out of the Total Retail Loss count entirely (ECR). Adoption of the broader model is real but partial: RILA’s survey found that 53% of respondents had fully or partly adopted the concept, and 47% were still thinking about using it the following year (RILA). Separately, 89% of retailers surveyed were aware of the framework and 64% of those believed it had made a difference in their organization (RILA). For a store manager, track your own shrink number for what the physical count shows, and raise anything outside that scope, such as vendor claims, as a separate line in the total retail loss conversation with corporate.
How Shrink Is Measured
Most retailers do not count shrink continuously. Shrinkage figures are usually generated through a physical inventory audit undertaken once or twice a year (ECR), with the gap between counts covered by estimates. Federal tax law allows this directly. Under 26 U.S.C. § 471, a method of determining inventories is not treated as failing to clearly reflect income solely because it uses estimates of inventory shrinkage “confirmed by a physical count only after the last day of the taxable year,” provided two conditions hold (26 U.S.C. § 471). First, “the taxpayer normally does a physical count of inventories at each location on a regular and consistent basis” (26 U.S.C. § 471). Second, “the taxpayer makes proper adjustments to such inventories and to its estimating methods to the extent such estimates are greater than or less than the actual shrinkage” (26 U.S.C. § 471).
NRF’s National Retail Security Survey measures shrink as a percentage of sales, covering losses from internal and external theft, operational mistakes and systemic errors (NRF National Retail Security Survey), and reports its own figures calculated at retail (NRF National Retail Security Survey). Sales is the standard denominator, but valuation basis is not standardized: authors cited by ECR suggest most retailers calculate shrinkage at retail prices, with between 20% and 40% using cost price or a mix of cost and retail (ECR), and ECR itself recommends valuing stock and asset losses at cost price unless a category definition says otherwise (ECR). Scope varies too: more than three-quarters (78%) of NRSS respondents exclude ecommerce goods loss from their shrink calculation, and 57% exclude supply chain loss or theft entirely (NRF National Retail Security Survey).
ECR’s comparison of two food retailers shows how much scope matters: including certain operational failures in the shrinkage calculation more than doubled the measured rate at one retailer compared with the other, which excluded them (ECR). Two retailers in the same category can report very different shrink rates simply because of what each one counts. Before comparing your number to another retailer’s or to a published benchmark, confirm the denominator, the valuation basis and what is in scope.
The Shrink-Rate Formula
Book inventory minus physical inventory is the dollar shrink figure. ECR’s Total Retail Loss report cites the widely used definition from Sennewald and Christman: shrink is the difference between book inventory, what the records show, and actual physical inventory, what a count on the floor finds (ECR). Divide that dollar gap by net sales for the same period to get the rate.
| Line | Item | How to get it |
|---|---|---|
| A | Book inventory | Pulled from your inventory system as of the count date |
| B | Physical inventory | Counted on the floor, same valuation basis as A |
| C | Shrink (dollar) | C = A − B |
| D | Net sales | Total sales for the same period as A and B |
| Rate | Shrink rate | Rate = C ÷ D |
If B comes in higher than A, the formula still runs: the result is a negative shrink figure, commonly called an overage, and it gets reported the same way. The formula only works if A and B are valued on the same basis, retail or cost, and if D covers the identical period. Run your own A through D through the retail shrink calculator, then write the valuation basis next to the result before you share it with anyone else.
Types of Shrink and Their Shares
NRF’s National Retail Security Survey breaks 2023 respondent shrink into five categories. External theft, defined as “shoplifting, burglary or break-in, robbery, credit card fraud and other retail crimes committed by non-employees,” averaged 36% of total loss (NRF National Retail Security Survey). Internal theft, defined as “acts of theft committed by an employee, or in collusion with an employee by outsiders,” came in at 29% (NRF National Retail Security Survey). Process, control failures and errors, covering cashier errors, inaccurate system pricing, accidental loss or damages and other operational or systemic errors, accounted for 27% (NRF National Retail Security Survey). Unknown losses, those not properly categorized, were 6%. Other, including expired goods and vendor or third-party losses, was 1% (NRF National Retail Security Survey). Damage sits inside Process. Vendor-related loss sits inside Other, with no line of its own. Theft overall, internal and external combined, accounts for nearly two-thirds (65%) of retailers’ shrink, and more than 70% in some sectors (NRF National Retail Security Survey).
NRF’s survey adds that external theft has surpassed internal theft as the largest contributor to overall retail loss, but internal theft remains the second largest, at an average of $2,180 per investigation, in line with 2021 and 2020 levels (NRF National Retail Security Survey). Treat these shares as estimates, not ledger entries. ECR’s Total Retail Loss report warns directly against guessing what portion of unknown loss belongs to a particular cause and then folding that guess into known-loss totals (ECR). The NRSS shares are self-reported by survey respondents, and the Unknown category exists precisely because some loss resists classification. A pie chart built from those shares reports what survey respondents believe happened to their inventory, based on their own estimates.
The shares also shift depending on who is doing the describing. Retail Dive’s account of NRF’s data states that 63% of lost inventory has internal causes while 36% comes from external theft, including shoplifting, credit card fraud, organized retail crime, cargo theft and other retail crimes (Retail Dive). That two-bucket framing does not line up category for category with the breakdown above, because NRSS reports internal theft, process and control failures, unknown and other as separate lines. Those four lines (29%, 27%, 6% and 1%) add up to 63%, the same figure Retail Dive reports as internal causes, so the two framings reconcile only if every non-external line counts as internal. The table below puts the two framings side by side.
| Framing | Category | Share of shrink |
|---|---|---|
| Two-bucket (Retail Dive) | Internal causes | 63% |
| Two-bucket (Retail Dive) | External theft | 36% |
| Five-category (NRSS) | External theft | 36% |
| Five-category (NRSS) | Internal theft | 29% |
| Five-category (NRSS) | Process, control failures and errors | 27% |
| Five-category (NRSS) | Unknown | 6% |
| Five-category (NRSS) | Other | 1% |
Before you quote a shrink-cause percentage in a huddle or a report, name which framing produced it, two-bucket or five-category, so the next reader does not assume employee theft alone explains 63 points of loss. Track internal theft on its own line, apart from ordinary operational error.
The Benchmark: 1.6% and $112.1 Billion
NRF’s National Retail Security Survey found that the average shrink rate rose to 1.6% in FY2022 from 1.4% in FY2021, representing $112.1 billion in losses across the industry (NRF). That figure is an average. The same survey’s table puts the FY2022 median at 1.4% (NRF), so half of responding retailers ran at or below 1.4%, and a smaller group of high-shrink retailers pulls the average up. Treat 1.6% as an industry-wide mean, something to compare against, rather than a target for any single store.
The FY2022 average is still the industry figure stores get compared against. In 2024, NRF said it would not release its annual shrink report that year, after conducting the research for more than three decades (Retail Dive), and NRF’s own spokesperson confirmed that the most recent available shrink statistics are found in the prior year’s National Retail Security Survey (Retail Dive).
NRF has also started describing shrink itself differently. NRF’s David Johnston said shrink is a lagging indicator, useful for describing what happened, but not for identifying where losses are building, where controls are weak, or for preventing loss before it reaches the profit and loss statement (NRF). NRF’s 2026 survey tracks incidents instead: retailers experienced a 12.4% decrease in shoplifting incidents and an 8.1% decline in retail merchandise theft in 2025 compared with 2024 (NRF). That trend describes incident counts. Cite 1.6% as the FY2022 benchmark it is, and pair it with your own count data for the number that reflects your store today.
Shrink by Store Type
NRF’s National Retail Security Survey found that shrink rates vary by retail sector. Pharmacy, grocery, department stores and mass merchandise average shrink rates over 2% of sales, while specialty apparel averages 1.9% (NRF National Retail Security Survey). Jewelry, watches, home furnishings and furniture, and footwear average 1.5% or less (NRF National Retail Security Survey).
| Sector band | Approximate shrink rate |
|---|---|
| Pharmacy, grocery, department stores, mass merchandise | Over 2% |
| Specialty apparel | 1.9% |
| Jewelry, watches, home furnishings, furniture, footwear | 1.5% or less |
The question practitioners actually ask, what counts as a good shrink rate for my store, does not have a published store-level answer. NRSS reports sector averages across many respondents, not a benchmark broken out by individual store format within a sector, so a grocery chain comparing itself against the sector’s “over 2%” average (NRF National Retail Security Survey) is comparing against a band, not a target. A single high-volume, high-shrink location can pull a chain-wide average above where most of its stores actually sit. Use the band for your own format as a rough compass, then build a store-level baseline from your last two physical counts rather than importing the sector number as a goal.
Grocery and Perishables
Food retail carries its own shrink math. FMI’s Food Retail Industry Speaks report puts 2022 food retail overall shrink at 2.7% (FMI), above the FY2022 all-retail average. USDA’s Economic Research Service treats shrink in food retail differently than NRSS does: it is a proxy for food loss in handling and display that also captures some unknown amount of theft, accounting error and other factors (USDA ERS), rather than a single theft-weighted percentage.
The item-level detail shows why perishables behave differently from dry grocery. USDA’s Economic Research Service found that average shrink for fresh meat, poultry and seafood was 12.7% in 2011-12, almost three times higher than the 2005-06 average of 4.5% (USDA ERS). At the individual item level, average annual supermarket shrink rates for fresh fruits, vegetables, meat, poultry and seafood ranged from 2.2% for sweet corn to 62.9% for turnip greens (USDA ERS), and a separate USDA chart put the average supermarket loss rate at 11.6% across 31 fresh vegetables (USDA ERS). The gap between sweet corn’s rate and turnip greens’ rate stretches plausibility as a theft pattern; it reads as a shelf-life and handling story. USDA notes its estimates come from a convenience sample and are not nationally representative (USDA ERS). Use the item-level figures to decide where to focus rotation checks, fast-spoiling greens first, rather than as a shrink target for the produce department.
What a Store Controls
A few variables stay inside a store’s control regardless of what NRSS, FMI or USDA report nationally. The first is the unit: report shrink as a percentage of sales, not inventory value, so the number matches NRF’s methodology (NRF National Retail Security Survey). The second is what you count: note what is excluded (ecommerce, supply chain, damages) before comparing your figure against a published benchmark.
Some drivers sit outside the store entirely. In NRF’s survey, 72% of respondents reported an increase in average value per incident in areas that increased felony thresholds (NRF). A manager cannot change a threshold, but can make sure those incidents are logged and counted.
Who may stop a shoplifter is a policy the store sets for itself. In NRF’s 2023 survey, 41.4% of retailers said no employees are authorized to stop or apprehend shoplifters, up from 37.9% in the 2022 survey; 58.6% named loss prevention or asset protection personnel, and 12.1% named non-LP personnel such as store managers, down from 19% (NRF National Retail Security Survey). Whatever a store decides, associates should know the rule, and the incidents that happen should still be logged and counted.
Put the denominator and the valuation basis on the same line as the rate before that figure leaves the store. A deeper total retail loss conversation with corporate, or a closer look at internal theft and external theft trends, both start from that same habit. Store leaders who want to see how larger chains handle this same discipline can look at what AP leaders at Belk, Big Lots and REI said about Total Retail Loss buy-in, and programs pairing this with frontline practice can start with what employee theft awareness training should actually cover.
What to Do This Week
- Pull your last physical count and confirm whether it was valued at retail or at cost, then write that basis next to the number.
- Separate process and control failures (cashier errors, pricing mistakes, damages) from theft in your next shrink report instead of combining them.
- Run your own book and physical figures through the retail shrink calculator to get a rate you can defend in a meeting.
- State your store’s format (grocery, pharmacy, apparel, mass merchandise) before comparing your rate to any published sector band.
- Flag any cause percentage you report as an estimate, not a confirmed figure, until a count backs it up.
Frequently asked questions
Does retail shrink include employee theft, or only shoplifting?
Internal theft is its own line inside shrink. NRF's survey defines it as theft by an employee, or by an outsider colluding with one (NRF), and the 2023 edition reported an average loss of $2,180 per internal theft investigation (NRF).
Is NRF's 2026 theft report the same thing as its shrink survey?
No. The 2026 Impact of Retail Theft & Violence survey polled senior loss prevention and security executives from February to April 2026, a different panel asking a different question from the shrink survey (NRF). It reports year-over-year changes in incident counts, which measure how often theft happens, not how much inventory a count finds missing.
Does locking up merchandise change a store's shrink rate?
It can shift the mix within shrink. NRF's survey notes some retailers report external losses above 40% of total shrink, while sellers who keep items such as personal care or luxury goods under lock and key tend to see a lower share of external loss (NRF).
Besides shrink itself, what else are retailers investing in to curb theft?
Budgets extend beyond inventory counts. NRF's 2026 survey found retailers investing in management training (72%), employee training (65%) and risk-intelligence technologies (61%) as part of broader theft and violence prevention efforts (NRF). Those programs target incidents; the shrink rate itself still comes only from comparing records with a count.