The Unit History a Manager Cannot Keep and You Can
Why this matters
On a managed property, everybody's memory is broken except yours. Tenants rotate every year or two. Property managers change companies faster than that. Management companies change entirely, and when they do the old work orders usually do not follow. Owners buy and sell. The only party with an unbroken record of what has actually happened to a specific piece of equipment in a specific unit is the shop that keeps going back to it. That continuity is worth real money in three specific arguments, and almost every shop throws it away by recording work orders as billing events instead of as history.
The four breaks in everyone else's record
Understand why the gap exists, because it tells you exactly where your record becomes valuable.
- Tenant turnover. The person describing the symptom has often lived there less time than the fault has existed. "It has always done this" means eighteen months, not eight years.
- Manager turnover. Individual property managers move between companies routinely, and each one carries their knowledge of a building out the door.
- Management company transitions. This is the big one. When a property changes management, the incoming company gets the leases and the financials. Historical maintenance work orders frequently do not transfer at all, and where they do, they arrive as a document dump nobody indexes. A property that changed hands two years ago has, in practice, two years of maintenance history in its current system.
- Ownership changes. A new owner inherits a building with no operating memory and immediately faces capital decisions about equipment they have never seen.
Each break resets somebody else's knowledge to zero. None of them touch yours.
A unit history is not a property file
Keep the two straight, because they answer different questions and decay at different rates. A property file answers how do I work here - access, isolation points, contacts, approval limits. A unit history answers what has happened to this equipment and what does that mean. The file is a current-state document you keep correct. The history is an append-only record you never edit, only add to.
The minimum event record
Nine fields per event. Every one of them exists to answer a question somebody will eventually ask you.
| Field | Why it is there |
|---|---|
| Date | Anchors everything else |
| Unit and equipment identifier | Ties the event to a physical thing, not to an address |
| Reported by, and what they reported | The symptom in the reporter's words, which is how the next call will also arrive |
| What was found | The condition, not the task. This is the field that wins arguments |
| What was done | The work performed |
| Labor hours | The only quantity that makes trend analysis possible |
| Part category | Generic category, not a part number, so patterns survive supplier changes |
| Callback linkage | Whether this event is a return on a prior event, and which one |
| Technician | So you can ask a human what they remember |
Callback linkage is the field shops skip and the one that changes conclusions. Without it, nine events on one unit could be nine distinct faults or three faults each visited three times, and those two situations point to opposite recommendations. One says the equipment is failing. The other says your diagnosis is failing.
What was found is the second one people skip. A record that says "replaced component" and nothing else is worthless in a replace-or-repair argument two years later. "Found scale accumulation restricting flow, component failed secondary to it" is the sentence that carries the case.
Where a finding is an active hazard, the record is the last step and not the first. Evidence of combustion spillage or a compromised heat exchanger means the appliance is shut off and tagged out of service on that visit, with written notice to the manager the same day. You do not leave it running and note it in the history for later.
Use one: winning the repair-or-replace argument with evidence
State a rule and hold to it, because a recommendation without a rule is an opinion and a manager cannot take an opinion to an owner.
The gate. Recommend replacement when, over a trailing 24 months, a piece of equipment has had 3 or more distinct repair events AND cumulative repair labor of at least 50% of the labor hours a replacement would take. Unit of analysis is the individual piece of equipment, not the dwelling unit and not the property. The window is trailing 24 months rather than lifetime, because acceleration is what justifies replacement and accumulation is not.
The case. Unit 7C's heating equipment, 45 months of history:
| Period | Repair events | Labor hours |
|---|---|---|
| Year 1 | 1 | 1.5 |
| Year 2 (single event, month 18) | 1 | 2.0 |
| Year 3 | 3 | 6.5 |
| Year 4, through month 45 | 4 | 9.0 |
| Lifetime total | 9 | 19.0 |
The trailing 24-month window runs from month 22 to month 45. The year-two event fell in month 18, outside it. So the window holds 7 of the 9 lifetime events and 15.5 of the 19.0 lifetime hours.
Run the gate. Replacement labor for this equipment type is about 8.0 hours, so the labor threshold is 4.0 hours. Events: 7, at or above 3, true. Labor: 15.5, at or above 4.0, true. Both conditions hold, so the AND gate is satisfied and the recommendation is replacement.
The contrast that makes the window matter. Unit 4A shows 5 lifetime events and 11.0 lifetime hours, which reads worse than several units that have never been flagged. But 4 of those 5 events fell in the first two years after a bad initial installation, and the trailing 24 months hold 1 event and 1.5 hours. Count: 1, below 3, false. The gate fails on the first condition and the recommendation is to leave it alone, regardless of how the lifetime totals look. A shop that runs the rule on lifetime numbers recommends replacing 4A, gets overruled by an owner who remembers the install problem, and loses credibility for the 7C recommendation that was actually correct.
Why the manager cannot do this themselves. Their system holds work orders from the current management company only, under whatever vendors were used. If the property transitioned at month 37, the manager can see months 37 through 45: 4 events and 9.0 hours. That is real but it is not a trend, and an owner can plausibly call it a bad year. Your record shows the same equipment ran 1 event a year for two years and then went to 3 and then to 4. That is a curve, and a curve is what makes a capital decision easy.
Use two: defending a charge nobody remembers
A new manager takes over and reviews the last year of spend on a property. They find two invoices from your shop on the same unit, fourteen months apart, and ask why they paid twice for the same repair.
Without a history, you have a work order that says what you did and no record of what you found, and the conversation ends with a credit you did not owe. With a history, the answer is a sentence: the first visit found one component failed with the rest of the system testing normal, and the second found a different component in a different part of the system, with the found-condition line and the callback-linkage field both saying so.
That is why callback linkage earns its place even when the answer is "no." An event explicitly recorded as not a callback is evidence. A blank field is not.
Use three: surviving a management handover
A management transition is when incumbent vendors get replaced, and it usually has nothing to do with quality. The new company arrives with its own vendor list and no reason to prefer you.
The play is specific and it has a deadline. Within two weeks of learning about a transition, send the incoming manager a per-property summary, unprompted: how many units you service, what is currently open, what is deferred and why, and which equipment is at or approaching your replacement gate. It costs a few hours across the portfolio.
What it does is reframe you. The new manager is walking into a building they know nothing about, under pressure to look competent to an owner who just hired them. You are the only person who can hand them a picture of the building. That is a materially different position from being the vendor the last company used.
Keep it strictly factual. A summary that criticizes the previous manager, the previous vendor, or the owner's deferred spending becomes a liability the moment it is forwarded, which it will be.
What to hand a manager: the one-page unit history
Not your database export. One page per unit, in their language:
- Equipment identified the way the manager identifies it
- A dated event list, one line each: what was reported, what was found, what was done, hours
- A current-condition line in plain language
- A recommendation, with the rule that produced it stated in one sentence
That last item is what separates a document a manager can use from one they cannot. "Seven repair events in 24 months and 15.5 hours of repair labor against roughly 8.0 hours to replace, which meets our replacement threshold" is a sentence a manager can read to a board or forward to an owner without adding anything. "This unit is on its last legs" is a sentence they have to defend themselves.
How to verify your history is real
The ten-minute test. Pick three units at random that have three or more events and try to produce a one-page history for each. If it takes longer than ten minutes, the data exists but is not retrievable, and unretrievable data loses arguments exactly as reliably as missing data.
Callback-linkage completion. Check what percentage of events have the field populated either way. Below about 80% and your trend numbers cannot be trusted, because the missing events could go in either direction and you have no way to know which.
Found-condition completion. Sample twenty events and count how many have a found-condition line that describes a condition rather than restating the task. This is usually the worst-performing field and it is the one that carries both the replace argument and the charge defense.
Continuity across a transition. Find a property that changed management and check whether your history spans the change. If your record restarted when theirs did, you have adopted their break, and the single most valuable thing you own on that property is gone.
References
- See related: How to Build a Property File Worth Having, The Multi-Property Work Order SOP
- See related: Building a Relationship With a Property Rather Than a Person, Reading a Commercial Building's Maintenance Log for Clues
- Trade-standard practice on service-history retention: keep equipment-level records for at least the expected service life of the equipment, since the record's value is highest in the replacement decision at end of life