The Laundry Contract Nine Years From Now
Why a property management company has to catalog paint colors, office chairs and water meter locations, and how a 10-year laundry contract became the reason we built our own system.
Some of our older properties have centralized laundry rooms. The laundry equipment is run by a vendor under a contract, and those contracts are ten years long. Buried in each one is an out: a window, somewhere between 180 and 270 days before expiration, when you can give notice. Miss the window and you are in for another term.
So the question is simple. How do you remember a 90-day window that opens nine years from now?
If I put every one of those dates in my Outlook calendar, I would overwhelm myself. If I put them on the property manager, they are probably not going to be here in nine and a half years. The three-ring binder in the office, with a site map and a hand-written note about the laundry contract, is not going to be found by the person who needs it.
That contract is why we built our own operating system. Not the whole reason, but the clearest one.
Data endpoints
The way I think about a property is as a set of data endpoints. Some of them are the obvious ones that come out of the property management software: leasing activity, expenses, the rent roll. Those get reported on every month and everybody in the industry has them.
Then there is everything physical about the building that nobody writes down.
Where do the paint swatches get saved, so when a unit needs touch-up the team orders the right color instead of color matching the last coat? Sixteen color matches later you have drifted to a different shade, one wall at a time.
Someone sits down in an office chair and it breaks. Where did that chair come from? There are no labels on furniture, if you have ever noticed. So we catalog every furnishing we buy, with the vendor and the model, and the next chair takes five minutes instead of a day.
Where are the water meters? In the old days there was a property binder with a site map and someone’s handwriting. Now everyone carries a phone with a GPS in it. A maintenance tech who has never set foot on a property, subbing for someone on vacation, should be able to open the app, tap one button, and have his map walk him to the shutoff. If that is not the easy button, I do not know what is. Without it, he can spend three hours looking, and by then the leak has damaged a lot more.
Four years ago, when Winter Storm Uri came through Texas, we were shutting off water right and left across the portfolio. I can tell you that button would have been nice to have.
Every endpoint is a scar
None of these endpoints came from a whiteboard. They came from something that went wrong.
Every one of them is an experience we lived through, after which I said: this is not going to happen again, we are going to have a better system. The laundry contract we missed. The paint that drifted. The chair. The water meter nobody could find in a freeze. I have hit my head on all of these, and I wrote each one down and solved it.
When I train people on the system, the reaction is usually the same. How did you think of this? The honest answer is that I did not think of it. I learned it. That is also why the list is never finished. I come up with new endpoints every week, because the properties keep teaching.
Capture is the project
When you describe any one of these things, it sounds simple. A contract date with an alert. A paint color on file. A pin on a map. Nobody hears it and thinks it is complicated.
The complicated part is that someone has to go get the data. The data capture is the project.
When we onboard a property now, we work through a checklist of roughly two hundred items that we need to be able to fly the plane. Contracts and their out windows. Meters and shutoffs. Finishes. Equipment. Vendors. Inspections and their cadence. Once all of that is loaded, the reminders and the escalations run themselves. Until it is loaded, the system is a nice idea.
We are in the loading phase. Some properties are far along and some are not, and the ones that are far along are the proof that it works. The property manager, whoever holds that seat in year nine, gets notified when the laundry window opens. If it is not acted on, it escalates. It flows into task management so that someone cannot say they handled it. You cannot close the task without completing the step.
Why we stopped buying it
There is a lot of proptech, and every product solves a fragment of the problem. Nobody has taken on the whole workflow, because the whole workflow is too big and the industry is too slow to change. You cannot turn the Titanic like a speedboat.
We were a victim of that for a while. We said yes to too many initiatives. Every bell and whistle you could bolt onto a property management company, we tried. Revenue management, automated market surveys, several flavors of AI. We failed at every one of them. Some of that was on us, because we over-implemented. Some of it is that you never really know a piece of software until you have signed. Have you ever met a software salesperson who told you the problems with their product?
So one summer I scrapped everything we could and started over with two problems. The first was leasing: the phone was ringing and nobody was answering, and the data underneath was dirty. The second was application fraud, which every operator in every city at every price point has felt. We fixed those two with outside tools and spent a year getting them right.
Then we turned to building our own system for everything else, because everything else is the endpoints, and nobody sells that.
What it is for
The point of capturing all of this is not the reports. It is that when you come into work, the system should know what you are supposed to be doing today, in priority order. And from the management side, we should be able to see which property and which person is on top of their game and which is behind, and use the best cases to coach and to train.
AI does not replace the humans in this business. Half our team are maintenance professionals fixing real things with their hands. What AI does is amplify the speed and quality of the output, and it can only do that if the data underneath is right. A model cannot find a water meter that nobody ever recorded.
Is it all built? No. The building blocks are in place, the loading is under way, and the laundry contract nine years from now will get its reminder. That is the standard: nothing on a property that someone has to remember, because someone will not.