The Three-Party Problem in Property Management Work

Why this matters

In retail residential work, one person calls you, decides, pays, and lives with the result. That single body is why most of a shop's process works without anyone having designed it. You tell the person standing in front of you what the repair takes, they say yes, you do it, they pay you before you leave.

Property management work splits that body apart. A tenant reports the fault. A property manager or site manager opens the work order and sets the priority. An owner, asset manager, or HOA board approves anything past a standing limit. An accounts payable clerk pays it on a cycle nobody in the conversation controls. Shops that lose money in this segment rarely lose it on price. They lose it on routing: a scope question asked of someone who cannot answer it, a yes taken from someone without the authority to give it, an invoice sent to a person who does not cut checks.

The four roles, and the bodies they live in

The roles are not people. They collapse and expand depending on the account. A hands-on landlord with three rentals is dispatcher, approver, and payer in one phone number, with only the reporter split off. A regional management company puts dispatcher and payer in the same office and leaves the approver as an absent owner who answers email on business days. An institutional owner gives you four distinct bodies plus a vendor platform that acts as a fifth gate, where an invoice can be rejected by software before a human ever reads it.

The split is not the problem, misrouting is

Nothing about the split is unreasonable. An owner who lives out of state genuinely should not be answering a 7am no-heat call, and a manager running four hundred doors genuinely cannot approve a roof without the owner. The split only costs you when a question lands on a role that does not own the answer.

Three misroutes cause most of the damage:

Scope from the reporter. The tenant says the landlord already agreed to replace the unit. They may honestly believe it. They are not the party who agreed, and they will not be the party who refuses to pay.

Authorization from the dispatcher, above their limit. The manager says go ahead because they want the tenant handled. When the invoice exceeds what they could approve, it goes to the owner cold, with no prior conversation, and the manager's incentive shifts from defending your work to explaining how it happened.

Delivery to the wrong endpoint. A complete, well-documented invoice emailed to the manager who dispatched the job is not in the payment system. It sits until someone forwards it, and your aging clock has been running the whole time.

Worked example: one no-heat order, run twice

A tenant in a managed multifamily building loses heat. Trip one, under a standing limit of 2.0 labor hours:

• 06:50 tenant calls the on-site manager.
• 08:15 the work order reaches your dispatcher.
• 10:30 your tech arrives after 0.4 hours of travel.
• 11:05 diagnosis complete. A failed control component. The repair needs a part plus 2.5 labor hours, so the order totals 3.0 sold hours counting the 0.5 diagnostic. That is above the manager's 2.0-hour standing authority.
• 11:10 tech calls the manager, who cannot approve it.
• 11:40 the manager emails the owner.
• 12:10 tech leaves for the next call with no answer.
• 15:20 the owner replies with a repair-versus-replace question.
• 15:35 the manager relays the answer: proceed with the repair.
• Next morning, a second mobilization: 0.4 hours travel, 2.5 hours on site.

Add up the technician clock: trip one is 0.4 travel plus 1.67 on site (10:30 to 12:10), which is 2.07 hours. Trip two is 0.4 travel plus 2.5 on site, which is 2.9 hours. Total consumed 4.97 hours, call it 5.0. Sold hours 3.0. So 60% of the technician clock this order consumed was sellable, on a job where nothing went wrong technically.

Now the same order with the split handled up front. The account carries a standing authority of 3.0 labor hours for diagnosis plus repair on a habitability item, and a named approver reachable by phone above that:

• 10:30 arrival, 11:05 diagnosis complete, 3.0 sold hours is inside the standing limit.
• 11:25 the manager confirms in writing and the tech proceeds. The part was a truck-stock item.
• 11:25 to 13:55 repair, depart 14:05.

Consumed clock is 0.4 travel plus 3.58 on site, about 4.0 hours, against the same 3.0 sold hours. Sellable share rose from 60% to 75%, and the tenant got heat the same day instead of the next.

Two things worth naming about that comparison. First, the technical work was identical in both runs, so none of the difference is skill. Second, sold hours and consumed hours are different currencies: the 3.0 is gross work you get to bill, and it still costs you technician time and travel to deliver, so what you actually keep is the margin on 3.0 hours. The point of the comparison is not a return figure, it is that the same sale cost you a full extra hour of clock because of where a question landed.

Where the split shows up in your numbers

You will see the cost in three places, and they are worth measuring per account rather than shop-wide, because one bad account can hide inside a good average.

Authorization latency. Clock time from diagnosis complete to a documented yes, counted only during the account's stated business hours. Use the median of the last 10 completed work orders for that account, not the mean, because a single board meeting will drag a mean by days and tell you nothing about the typical order.

Second mobilization rate. The share of orders in that same set of 10 that required a return trip caused by waiting on an approval, not by a part or by access.

Days to payment. Measured from the date the invoice hit the payer's endpoint, not the date you emailed the manager. If you measure from your own send date you are measuring your submission habits, not their cycle.

Here is the rule the first two support. If the median diagnosis-to-authorization latency across an account's last 10 completed orders exceeds 4 clock hours inside the account's stated business hours, the standing limit is set too low for the work that account actually sends you. Raise it in one step, to the labor-hour value that would have cleared 8 of those 10 orders without escalation, and leave the other 2 to escalate. Do not raise it to the value that clears all 10, because the largest order in any set of 10 is usually a genuine capital decision that an owner should be making.

Run the worked example through that rule. The order needed 3.0 sold hours against a 2.0-hour limit, so it escalated. If 8 of the last 10 orders for that building would have cleared at 3.0 hours and the remaining 2 were a full equipment replacement and a roof, the new limit is 3.0 hours, not the 6.5 the replacement needed.

What collapses the split and what widens it

The method above is the same everywhere. What changes is how far apart the roles sit, and that changes what you have to build before the first call.

• Single-family rental, hands-on landlord. Three roles collapse into one reachable person. Treat it close to retail work, with one exception that never relaxes: scope still does not come from the tenant.
• Small management company. Dispatcher and payer share an office; the approver is an absent owner on email. The binding constraint is business-hours reachability, so your standing limit matters most for evenings and weekends.
• HOA or condo association. The dispatcher is a community manager, the approver is a board on a monthly meeting cycle, and the governing documents may cap what the board itself can spend without a membership vote. This is the widest split, and the one where a routine repair can sit for weeks because it was quoted three days after a meeting.
• Institutional owner or REIT. Four separate bodies plus a vendor platform with its own compliance and invoicing rules. The platform is not a formality; it can reject an invoice for a missing field with no human involved.

Reading a new account for its split before you quote

Ask these five on the first serious conversation, and write the answers into the account record rather than into someone's memory:

1. Who opens a work order, and who closes it? The closer is who decides whether your work is done, which is not always the person who dispatched it.
2. What can you approve without asking anyone? Managers state this in money. Translate it into a labor-hour and parts test your tech can apply on site, because a tech on a roof cannot do a running total.
3. Who do I call at 9pm on a Sunday, and what can that person approve? The second half of that question is the one shops forget, and it is the half that matters.
4. Who is the tenant contact, and may we schedule directly with them? Some accounts require all tenant contact to route through the office. Assuming otherwise is a fast way to lose an account.
5. Where does the invoice go, and what has to be on it to pass first look? Ask for a work order number, a PO field, a required attachment list.

If a prospect cannot answer 2 and 3 on the spot, that is not a disqualifier. It is an open item that has to close before your first after-hours call, and the honest ones will tell you they need to check.

How to verify you got this right

Pull the last 10 work orders for one managed account and tag each with which role gave you the yes. If a tenant appears in that column even once, you have a live exposure on that account, not a past one, because the same habit will repeat.

Then count how many of those 10 needed a second trip purely to wait on an approval. Anything above 2 of 10 says the limit or the after-hours contact is wrong, not that the crew is slow.

Last, ask the manager to state your standing limit back to you from memory. If they cannot, it is written somewhere but it is not operative, and your tech will get a verbal yes that nobody upstream is bound by.

References

• See related: Who Actually Signs and Who Only Calls; How to Set Approval Limits With a Property Manager; The Property Manager as a Repeat Client
• See related: REIT and Institutional Property Management Vendor Onboarding; Rental Property Manager Work
• Trade-standard practice for work-order authorization and vendor invoicing on managed properties