How to Train Associates to Write Better Incident Reports?
How to Train Associates in Loss Prevention Report Writing?
Train associates and LP agents to write loss prevention reports in first person, naming exactly who saw the incident and whether it was live or on camera. Describe actions and merchandise with specific, verifiable detail, including facts that cut against the conclusion. Lock date, time and location as fixed fields, leave the narrative in the associate’s own words, and correct each new draft with live feedback.
Who reads an incident report
A report written only for a manager’s eyes reads differently than one that ends up in front of strangers who never set foot in the store. Adam Parker, director of loss prevention at Lamps Plus, raised this point on Episode 17 of the Loss Prevention Magazine Podcast (listen on Buzzsprout), produced by LP Magazine and aired April 13, 2020.
“The writer needs to understand how many different audiences will view the report. It may not be just an internal audience.”
A Georgia court admitted a store’s loss prevention report as a business record specifically because the store prepares one in every instance of shoplifting (Thompson v. State). In a false-imprisonment case, a Virginia court held that the store’s own documentary evidence, including an earlier incident report a manager had written, made probable cause a question for the jury (Dill v. Kroger). Train every associate on the assumption that a prosecutor or jury may read the same report the manager files tonight.
Writing in first person
Vague phrasing that hides who saw the incident leaves a prosecutor unsure which witness can testify to the theft. California’s POST training materials tell officers to write in first person specifically so the reader knows what the writer actually did and saw (POST LD 18), and a prosecutors’ evidence checklist for retail crimes says a report must document who the actual observer was (King County PAO). Passive phrasing also leaves unclear whether the view was live or on a monitor. Put the observer’s name and whether the view was live or on camera in a fixed field at the top of every incident report, ahead of the narrative.
Facts versus impressions
Calling a walk-out theft without saying which hand took which item from which shelf collapses as soon as a supervisor asks what the associate actually saw. California’s POST guidance says conclusions should always be accompanied by the supporting facts (POST LD 18), and that information which conflicts with a stated conclusion must also be included (POST LD 18). Fern Abbott, who directs training at AFI Security Training Institute, made the plain-language point on the same podcast.
“Write in plain English… Don’t use a $4 word when a 25¢ word will do. I’m not impressed that you know big words. And if I don’t know what that big word means, you’ve just alienated me.”
Swap ‘concealed merchandise’ for which hand, which shelf edge, and which SKU left the fixture. Train associates, as part of ongoing associate training, to describe the clothing, the item and where they stood in plain words, and leave out any label for what kind of person they think they saw.
Templates and fixed fields
A report template exists to make sure date, time, location and a case number always land in the same place. It should never dictate how the narrative is written.
“The abbreviation RP means reporting person, but not everyone knows what that means… Don’t acronym the crap out of a report.”
Troy Willmon, a former investigator with the Ellis County, Texas Office of Emergency Management, made that point on the same episode. A template can use RP as a field label, but the narrative should spell out “reporting person” for a reader who has never seen the form. The same push for plain specifics applies to merchandise: a prosecutors’ checklist for retail crimes requires descriptions that go beyond the boilerplate phrase “various items of merchandise” (King County PAO), and in Kroger v. Suberu a loss prevention employee stapled a printed list of the scanned items to his report, which gave a Texas court something concrete to test against the plaintiff’s claim (Kroger v. Suberu).
| Report element | Fixed field | Narrative |
|---|---|---|
| Date, time, store number | Yes | No |
| Reporting person’s name and role | Yes | No |
| Observer’s name, and whether the view was live or on camera | Yes | No |
| What the writer personally saw, in order | No | Yes |
| Item brand, color, size | Partial (SKU/UPC) | Yes (visual detail) |
| Suspect’s own statement, verbatim | No | Yes |
Build every template so the only free-text field is the narrative itself, and drop dropdown options like “suspicious behavior” that let a guess pass as a fact.
Narrative before camera footage
Have the associate finish their own description before anyone pulls the cameras. Research on report writers found that officers should write down their memories of an incident before watching body-worn camera footage, and mark any changes made after watching it (Vredeveldt et al., 2021). Keep the CCTV player closed until the associate finishes that paragraph. California’s POST guidance says that when footage is used to describe an incident, the report should state plainly that it reflects what the cameras captured (POST LD 18), so put whatever the footage confirmed or changed in a separate labeled note after the associate’s own paragraph.
Theft aggregation and report consistency
Florida law lets prosecutors aggregate thefts from more than one location within a 120-day period, and a combined total of $750 or more becomes a felony (Florida Stat. 812.015). California allows the value of property stolen across one or more thefts to be aggregated into a single count (California Penal Code 490.3). Both statutes depend on someone connecting an incident at one store to an incident at another, and there is no centralized reporting system for organized retail crime cases to do that linking automatically. Standardize item descriptions and reporting-person names across every store’s template, using a structure like the ORC observe-and-report form, so a report from one location can actually be matched against one from another later.
What prosecutors need from a report
Christopher Walden of the Utah Attorney General’s C.A.S.E. unit wants the case basically complete before it reaches law enforcement (LP Magazine). When a referral reaches Oklahoma City Police detective Brandi Streeter, she expects loss prevention reports, written statements, CCTV footage, and photographs attached to it (LP Magazine). Poor report writing, in former Sacramento County DA Anne Marie Schubert’s account, is a primary reason cases never prosecute (LP Magazine). Do not send the referral until the written statements, CCTV footage, photographs, and item list are already in the file.
Reporting requirements beyond the incident report
A shoplifting write-up is not the only record a store must keep after an incident. California’s SB 553 now requires every employer to record information in a violent incident log for every workplace violence incident (California Labor Code 6401.9), built from employee accounts, witness statements and investigation findings, without personal identifying information, plus the completer’s own name, job title and completion date. For the manager training this same law requires, see what California SB 553 requires a manager to train. The log has to be kept for at least five years (California Labor Code 6401.9). Federally, OSHA requires Form 301 within 7 calendar days after an employer learns of a recordable work-related injury, and its field asking what the employee was doing just before the incident specifically instructs the writer to be specific (OSHA Form 301). New York does not require retail employers to keep incident records, though the state’s labor department recommends it (NY DOL). File the violent incident log on its own form, keep OSHA Form 301 beside it for a recordable injury, and leave the shoplifting write-up as a third record.
What to do this week
- Rewrite your incident report template so every fixed field (date, time, location, reporting person, observer and live-or-camera, item SKU) sits above a blank narrative field labeled “what I saw, in my own words.”
- Pull five recent reports and circle every euphemism (“suspicious,” “acted nervous,” “various items”) and replace each with the specific behavior or item it was standing in for.
- Add one line to your training script: write your memory of the incident before pulling the camera footage, then label anything added after review.
- Walk one associate or LP agent through a live correction on their next report instead of a slide deck, then repeat it with the next report they write.
- If your store operates in California, confirm your violent incident log is a separate document from the shoplifting incident report, not the same form doing double duty.
Frequently asked questions
What should replace a phrase like 'various items of merchandise' in a theft report?
List each item by brand, size, color and SKU or UPC if the register captured one. The King County prosecutor's evidence checklist asks for item descriptions more specific than 'various items of merchandise' (evidence checklist); specific items let a court verify the loss independently.
Does a workplace violence log replace the shoplifting incident report?
No, they serve different purposes. California's SB 553 requires a violent incident log for workplace violence, built from witness and investigation input, retained for a minimum of five years (the SB 553 statute); keep it separate from the shoplifting report, which documents theft, not violence, and follows its own template.
Who should train associates and LP agents to write reports, and how often?
Correct real drafts one on one rather than lecture from a slide deck; run it during a regular huddle or microlearning session, then repeat with each associate's next actual report so the correction sticks to real writing, not a hypothetical example.