How to Build a Property File Worth Having

Why this matters

Every shop that works managed properties has property files. Most of them are an address, a manager's phone number, and a note somebody wrote in 2019. The test of a property file is not whether it exists, it is whether a technician who has never been to that building can take a call at ten at night and start work without phoning anyone. Almost no file passes that test, and the cost shows up as unbilled phone time on every call, second visits for information that was knowable, and a shop where two people carry the whole portfolio in their heads.

Step 1: Set the file's job as a test, not a template

Write the test at the top of the file and build backwards from it: a tech who has never been here can get in, find what they need, know who to call, and know what they are allowed to spend, without calling the office.

That test decides every inclusion argument you will have. Anything that helps a stranger at night goes in. Anything that only helps the person who already knows the property does not. Templates invite you to fill fields; a test tells you which fields matter.

Skip this and the file becomes a collection of everything anyone ever recorded, which is the same as having nothing, because at ten at night nobody scrolls.

Step 2: Build three layers, and keep them separate

Keeping them separate matters because they go stale at completely different rates and they have different owners. The commercial layer changes at renewal. The human layer changes when a manager moves companies, which is often. The physical layer changes every time equipment is replaced. A single blended page ages to the speed of its fastest-changing field, which means the whole thing feels untrustworthy and gets ignored.

Step 3: Populate it from the first three visits, not from a survey

Nobody funds a site survey, and a survey done before you know the property records the wrong things anyway. Instead, add roughly 20 to 30 minutes to each of the first three visits at a new property, with a specific assignment each time:

1. First visit: access and the human layer. How you got in, who let you in, what the code or key arrangement was, who answered the phone, who did not.
2. Second visit: isolation and infrastructure. Main water shutoff, gas meter, electrical panels and main disconnect, unit isolation valves, mechanical room location, any fire-system control valve.
3. Third visit: equipment inventory for the units you touched, plus a walk of anything shared.

Three visits gets you a file built out of real experience rather than an intention. It also means the file is written by the people who will use it, which is the only reliable way to keep it honest.

Step 4: Get the access section right, because it is the one that decays

Access is the most-used and fastest-rotting part of the file. It needs six things, and each one has a failure mode when it is missing:

• The method. Lockbox, key at the office, tenant present, manager meets you, code at the door. Missing means the tech arrives and improvises.
• Where the code or key physically is. A code with no location is half an answer.
• Who holds the backup. Because the first method fails often enough to plan for.
• Occupied or vacant, per unit. It changes the entire visit, from notice requirements to how long you should expect to wait.
• Parking, loading, and after-hours building entry. Underground garages, gated lots, and locked lobbies are the reason a tech is standing outside at ten at night with a valid unit key.
• The last date this was verified. Without it nobody knows whether to trust the rest.

Handle codes with some care. Store them in the system with visibility limited to dispatch and the assigned technician, and keep them out of group chats and off printed route sheets that live in trucks. A property manager who learns their codes are circulating in a shop's group thread will not give you a second one.

Step 5: Record isolation points, and know the one valve you do not close

This is the section that pays for the whole file, because it is what a tech needs in the first ninety seconds of an emergency, when there is nobody to ask.

Record the main water shutoff and how it operates, the unit or riser isolation valves and what each one actually serves (verified, not assumed from a label), the gas meter and its shutoff, the electrical panel serving each unit, the main disconnect, and the mechanical room access.

Two hazard notes belong in the file itself, not just in training, because the file is what gets read at the moment:

Gas. On a gas odor in a unit, corridor, or common area: everyone leaves the building immediately, no light switches, thermostats, or appliances touched, no phone used inside, call the gas utility emergency line from outside, notify the manager after, and do not re-enter until the utility clears it. The meter location goes in the file for the utility's benefit and for post-incident work, not as an invitation to go operate it during an active odor call.

Fire protection. If isolating water would close a valve serving a fire sprinkler system, do not close it on your own authority. Taking a water-based fire protection system out of service is an impairment, and NFPA 25 requires the system owner to run an impairment procedure, including notification, before and during the outage. Mark those valves in the file explicitly, by tag or location, so a tech chasing a leak at night does not close one because it was the nearest handle.

Step 6: Keep the equipment inventory to a minimum viable field set

An inventory nobody can complete is an inventory that stays empty. Six fields per piece of equipment, all of which a tech can capture from a nameplate photo in under a minute:

• Unit or location served
• Equipment type and capacity from the nameplate
• Age or install date, or the best evidence you have and its source
• Serial or asset identifier
• Last work you did on it, with the date
• Anything nonstandard: a previous repair, an access difficulty, a part that is not the usual one

The age field is the one people fudge. Record what the evidence supports and label it: "nameplate date," "manager's records," or "estimated from the install of the adjacent unit." An estimated date honestly labeled is useful. An estimated date presented as fact will be quoted back to an owner in a replace-or-repair argument, and being wrong in that room costs more than not knowing.

Step 7: Set staleness rules with triggers, not good intentions

Every field gets a re-verify rule, and the rules are different because the decay rates are different.

Then one escalation rule that catches the file drifting as a whole: if two or more fields on a property's file are found wrong during a single visit, the property gets a full re-verify pass rather than a spot fix, scheduled inside 30 days. Unit of analysis is the property, not the field. Two wrong fields is the signal that something changed at the property that nobody told you about, and spot-fixing the two you happened to hit leaves the rest wrong.

Step 8: Make it open on a phone in under fifteen seconds

The best file in the world is worth nothing if the tech has to call the office to get it. It must open from the dispatch record on a phone, and the access and isolation sections must be at the top, because those are what get read under pressure.

Test it the way it will be used: hand a phone to a tech who has never been to the property and ask them where the main water shutoff is. If it takes more than fifteen seconds, the file is organized for whoever wrote it rather than for whoever needs it.

Worked example: what the file was worth

A shop with six managed properties measured two things before building files, across a quarter of calls.

• Pre-work phone time: an average of 11 minutes per call spent calling the office, calling the manager, or hunting for a shutoff before work started.
• Second-visit rate: 22% of calls needed a return trip for a part or a piece of information that was knowable in advance.

They then ran the three-visit population from step 3: six properties, three visits each, 18 visits with roughly 25 extra minutes on each, which is 450 minutes or about 7.5 technician-hours. Add roughly 4 hours of office time to build the commercial and human layers. Total build cost about 11.5 hours, spread across a quarter so no day carried it.

Re-measured a quarter later:

• Pre-work phone time: 4 minutes per call, down 7 minutes.
• Second-visit rate: 13%, down 9 percentage points.

Across roughly 240 calls a year on those six properties:

• Phone time: 7 minutes saved on 240 calls is 1,680 minutes, about 28 hours a year.
• Second visits: 22% of 240 is 52.8, and 13% of 240 is 31.2, so 21.6 fewer return trips. At roughly 1.4 unbilled hours each for drive, re-access, and setup, that is about 30 hours a year.
• Total: about 58 unbilled hours a year recovered, against a one-time build of about 11.5 hours.

Be precise about what that 58 is, because it is easy to overclaim. Both sides of the comparison are the same currency: technician hours that produce nothing. The build cost 11.5 of them and the file returns about 58 of them a year. That is recovered capacity, not revenue. It becomes revenue only to the extent you have work to put into the freed hours, and on a shop with a thin board it may simply mean people go home on time, which is worth having but is not the same claim.

The second-visit number is the more interesting half anyway. A 22% return rate on a portfolio means roughly one call in five ends with a tenant still waiting and a manager fielding a second complaint. Cutting it to 13% is a service change the manager notices, and that is what gets an account renewed, not the hours.

How to verify the file is real

The stranger test, on a live call. Once a quarter, dispatch a tech who has not been to a property and ask afterward how many phone calls they had to make before starting work. The target is zero. One is acceptable. Three means the file is decorative.

The correction rate. Count corrections logged against each property's file per quarter. Zero corrections is not a good sign, it usually means nobody is opening it. A steady trickle of small corrections is what a live file looks like.

The re-verify trigger. Check whether any property hit the two-wrong-fields rule and did not get its full pass inside 30 days. That gap is where a file quietly reverts to being the address and a phone number from 2019.

References

• NFPA 25, Standard for the Inspection, Testing, and Maintenance of Water-Based Fire Protection Systems, including impairment procedures for systems taken out of service
• See related: Building a Relationship With a Property Rather Than a Person, The Multi-Property Work Order SOP
• See related: The Unit History a Manager Cannot Keep and You Can, The Wrong Unit Reported on a Multi-Unit Property