Editor's Note
Something crossed my desk this week that predicts which appliance breaks next. The prediction part works. This issue is about the layer underneath it, where most small operators will get stuck.
Predictive maintenance is here. Getting it to run is another matter.
ForVue scores every appliance in a portfolio each night and ranks what's likely to fail. HVAC, water heaters, electrical, plumbing. It runs on Weibull-Bayesian failure analysis, the same reliability math used in aerospace and industrial maintenance for fifty years.
Pricing starts around $2.50 per unit per month, and one tier is built specifically for landlords running 1 to 50 properties without a maintenance team. That is unusual. Most tools in this space are priced for portfolios ten times that size.
The integrations check out. Rent Manager confirmed a live production partnership in their own August bulletin, and Hospitable lists ForVue as a verified marketplace app.
Here's my caution: I could not find a single independent user review. Not on G2, not on Capterra, not in any operator forum. Everything written about this product comes from the company or its integration partners. That doesn't make it bad. It means nobody outside has publicly tested it yet.
What the model actually needs
The engine requires | Most small portfolios have |
|---|---|
Model and serial number | Maybe, if the closing folder survived |
Install year | Rarely written down |
Full repair history | Scattered across texts and invoices |
Usage intensity | Almost never tracked |
Prediction math is only as good as what feeds it. A failure model with no install date falls back to the average, which is roughly what a good contractor tells you for free.
The company knows this. They built a scanner that reads appliance data plates through a phone camera, pulling brand, model, serial, and install year automatically. That feature exists because typing four fields per appliance across twelve properties was never a realistic ask.
Where the real constraint sits
Emergency repairs cost several times what the same job costs when scheduled. Every operator knows this already. It keeps happening because nobody can see which unit is about to go.
So the hard part was never the prediction. It's the record-keeping, and record-keeping is dull work that nobody wants to spend a weekend on.
There's a test that takes five minutes. Pick your worst property. Can you name the install year of its water heater? If yes, predictive tools have something to work with. If no, that's your actual starting point, and it costs nothing but an afternoon with a phone camera.
When to ignore all of this
If you manage under five units and know every appliance personally, you already are the prediction engine. This becomes useful somewhere past the point where you stop remembering which unit has the old furnace.
Did You Know?
In Buildium and NARPM's 2026 State of the Property Management Industry Report, a survey of over 3,200 professionals, 93% of property management companies said expenses rose over the past year. Vendors, materials, and insurance led the increase. Preventing one emergency call matters more in that environment than it did three years ago.
Until next week,
Tomas