Report 209 · Luxury Home Security
Researchers in Austria sat down with convicted burglars in seven prisons and asked how they plan a job. The most useful answer for anyone with an alarm system: some of them test it first. They start a break-in, walk away, and watch whether anyone comes. If nobody does, they come back.
Much of what gets repeated about how burglars think traces back to one American survey of 422 offenders, published in 2012, which I took apart in Report 160. So a newer study from a different country is worth reading closely. In October 2025 the Institut für Konfliktforschung in Vienna published "Modus Operandi," a report by Günter Stummvoll (IKF) and Patricia Rosenauer of the Austrian safety board KfV (Kuratorium für Verkehrssicherheit), which commissioned it. The report is in German. I read it in full, and the translations below are mine.
The core of the study is 35 one-on-one interviews in seven Austrian prisons, run in December 2024 and January 2025. Three were broken off, leaving 32 usable interviews (24 men, 8 women), each 15 to 45 minutes and recorded. The team also coded 30 court files from the same prisons and interviewed specialists from the Vienna police.
Two limits up front. First, these are people who got caught, talking about themselves, which is a known bias in every offender study. Second, "burglary" here is broad: only half of the interviewees said explicitly that they had broken into homes. The rest described cellars, offices, restaurants, schools, kiosks, hotels and pharmacies. The authors also attach an unusual warning to their own chapter on technique, saying it reads like a burglary manual and should be handled carefully when published. I've kept this report at the level a homeowner needs.
Here is the passage that matters most for an alarm owner. Some offenders, the authors write, test whether an alarm system works by only starting a break-in, for example forcing just the terrace door. Then they move away for a few minutes and watch whether an alarm went off or police were called to the scene. If neither happens, the break-in continues.
A different interviewee, who broke into pharmacies, described the same logic: fake break-ins in which only a window is smashed, to see how the alarm chain plays out. In his account, the police took between 5 and 15 minutes. That's one offender's claim, not a measured response time, but it shows what he was timing.
Read that as an engineer and the lesson is uncomfortable. To this kind of offender, the alarm isn't the deterrent. The response is. A siren that sounds and stops with nobody arriving teaches the burglar that the house is safe to enter.
The report also says offenders work out how an alarm functions before going in. Its example: window alarms commonly use a magnetic contact that trips when the frame opens, so the offender may break the glass instead of opening the window. The noise means they pick a secluded house for it. Translation for a homeowner: a contact sensor watches the frame, not the pane, and on its own it leaves the glass uncovered.
On cameras, the interviews are mixed. One offender said he keeps his eyes down during break-ins so his face doesn't show on video, and that he expects a live feed to the owner, who then calls the police. Others, talking about cellar break-ins, didn't find video surveillance much of a deterrent. One put it in four words: "Bis da jemand kommt..." (roughly, "By the time someone comes..."). Again, the camera matters to the extent someone acts on what it sees.
In the 30 court files, the tool most often recorded was simple prying or forcing (18 of 52 mentions). Electronic jammers appeared in 4 percent of mentions, a small share, but present. I covered what a jammer can and can't do to a wireless alarm in Report 197.
The study's summary describes a small group of specialists who go after high-security targets, naming villas, car dealers and jewelers, with considerable technical and electronic effort, for money and jewelry worth several hundred thousand euros, sometimes organized in international networks. The interviews give a sense of their arithmetic. One offender said a haul had to be worth at least 10,000 to 15,000 euros to justify the risk. Another, describing professional crews, said he wouldn't pick up the phone for 1,000 euros, and that contacts in the Balkans were only called in when 400,000 euros or more was on offer.
How they find out when a house is empty is mostly ordinary. The report lists holiday plans mentioned when booking tradespeople, posts on social media, and hiding places discovered by delivery and construction workers. Police added cars normally parked in view (their absence shows nobody's home), hedges that hide an intruder once inside, and homes in well-off areas with no alarm at all. Key lockboxes left at entrances for care services were simply ripped off and opened.
Time is the other variable. A typical apartment break-in, the report says, is done in five to ten minutes. If the offender has seen that a house is empty for the weekend, it can stretch to half an hour.
Most of the report's recommendations are habits: don't announce absences, never leave windows or terrace doors tilted open even briefly, lock up every time, close the garage door fully, maintain the alarm regularly, keep little cash at home, and know your neighbors, because offenders read neighborly contact as social control. Police experts added one I'd underline for large homes: arm the perimeter alarm even when you're home, since burglars don't always know whether someone is in. And if you come home to a break-in, don't go inside; the burglar may still be there. The report finds that burglars generally want to avoid any confrontation, and the police advice is to let them run and remember details rather than block the exit.
My reading, from designing these systems: the "test" works on any alarm whose only output is noise. What defeats it is a system that turns a single forced door into a verified event someone acts on, whether that's a monitoring center that confirms with video and dispatches, or a private response, plus layered detection so the glass is covered as well as the frame. A partial break-in that ends with nobody coming should itself raise a flag, because the study suggests it may be a rehearsal. I've written about why police may not respond to an unverified alarm in Report 072.
I help design the AI security systems for a veteran-owned (SDVOSB) home-security company run by fellow veterans. I do not own that company and earn nothing from this link. Full policy here.
How common the alarm test is. The report describes it from interviews and gives no count. It's a reported tactic, not a measured rate.
The 5 to 15 minute police time. One offender's account, not police data.
Transfer to other countries. This is an Austrian sample of mostly urban, often addiction-driven offenders. The professional crews that target estates are described mainly through the police interviews and a few offenders.
A 2025 Austrian study of 32 imprisoned burglars found that some offenders treat an alarm as a question, not an answer: they trip it on purpose, wait, and come back if no one responds. Contact sensors that watch only the frame can be sidestepped by breaking glass, and cameras deter mainly when someone is watching. For a high-value home, buy the response and the verification, not just the siren, and treat an aborted break-in as a warning.
U.S. Army combat veteran (Counter-IED / Electronic Warfare), peer-reviewed researcher in microwave spectroscopy, and founder & CEO of Shroombiosis. Consults on laboratory operations, AI, and supplement formulation.