Why I thought this would be easy
Monday morning, June 12, 2023. 7:40 a.m. I'm standing in the entry hall of a 12-unit apartment building we'd just finished renovating. The air smells like fresh paint and, faintly, burnt dust. The smell is courtesy of an iRobot Roomba 877 vacuum cleaning robot doing its first run.
I'd told our on-site manager the day before: "It's a robot vacuum. Just let it work."
It didn't just work. Four interruptions in twenty minutes, all at the same carpet-to-tile transition. And by the time it limped through two rooms, the bin was full. Not the robot's fault. Mine. I'd read "charging dock" on the spec sheet and mentally filled in "self-emptying dock." Those are two different things. On a 12-unit property, that difference meant someone had to empty the bin every couple of rooms—or the floors looked worse than before we started.
That was the first mistake. It was not the last.
For context: I've handled facilities and procurement for a property management group for nine years. Sixty-ish residential units and a handful of small commercial spaces across three buildings. My job is to watch the numbers and the timelines, and to keep the on-site teams from having to reinvent processes every time we buy something new. By June 2023, I'd caught enough issues to think I knew what "plug-and-play" actually meant.
That confidence cost me roughly $4,200 over six months. Not one dramatic loss—a slow leak. Some of it was literal spend. Most of it was the hours my team spent re-training, re-mapping, re-cleaning, and re-doing what we'd thought was done.
Four purchases, four assumptions, four corrections. Here's the honest version, because the pattern matters more than any single device.
The four purchases (and the mistake behind each)
1. iRobot Roomba 877 vacuum cleaning robot. I bought four of them for two small office spaces and the shared lobby. Total: about $1,200. My reasoning: they're workhorses, they're reliable, and the price was right. My mistake: I sized them like I'd size a home vacuum. The 877's standard bin is around 0.6L. Fine for a living room. Not fine for a lobby with 40+ people a day. The auto-empty problem was the small issue. The bigger issue was that I'd bought a fleet before piloting one. Two units with a smarter routing schedule would've covered both spaces. I bought four—four bins to empty, four firmware sets to maintain, four docks to place. Simpler isn't always cheaper.
2. The iRobot Combo j9+ review I read—and misread. When I skimmed the 2024–2025 iRobot Combo j9+ review coverage, I thought I'd found the upgrade path. The reviews focused on home use, which made sense—the product is designed for homes. What I glossed over: the mapping time, the transition adapters, the mop-pad maintenance, and the fact that "combo" means "vacuums and mops on a schedule you configure," not "vacuums and mops everything, forever." I tested one in the lobby. It worked well. I then rolled it out to a space where the cleaner had assumed "it handles everything." It doesn't handle everything. It handles what you tell it. We rebuilt the map twice and lost about six hours of onboarding. A 30-minute pilot with a printed "here's what it does and doesn't do" sheet would have avoided both.
3. UltraCool CP70 portable evaporative cooler. This was the expensive one. One of our buildings has a warehouse-adjacent lounge that gets hot in summer. I ordered the CP70 portable evaporative cooler after reading the spec sheet. On paper, it looked like the right tool. What the spec sheet did not say—what the vendor who took my order did not say—is that evaporative coolers raise indoor humidity. In my climate, in that room, that meant the lounge went from "hot" to "hot and muggy," and we started seeing mildew on the window trim within three weeks. The unit was returned. The trim wasn't. Cleaning and resealing ran me $890, plus a replacement dehumidifier at $240. The vendor who eventually helped me—a different dealer—spent ten minutes on the phone asking about square footage, ceiling height, and outdoor humidity before recommending anything. That ten minutes was worth more than the entire CP70 transaction.
4. Carbon monoxide detector for the home. I'll be honest: I didn't think about this one until a tenant asked. We have gas water heaters in two of the buildings. I ordered CO detectors for the shared mechanical rooms and the units closest to the heaters. My initial approach: "Install them, they're good for ten years." Then I read the actual manual, which—lesson learned—is not optional reading. Ten-year sensor life is typical. Monthly test-button checks are not. We caught a failing detector six weeks after install. Manufacturing defect, likely rare, but the only reason we caught it was because the on-site manager happened to press the test button during a routine walk-through. Had we relied on "it's new, it's fine," we might still be relying on a dud.
5. The gas water heater diagram mistake (a free lesson, almost). The pilot on one of the water heaters was flickering. I pulled the manual, found the "how does a gas water heater work diagram," and decided I could diagnose it. The diagram was clear. My reading of it was not. I spent $145 on a service call for what turned out to be a 30-second thermocouple reset. The tech was polite. He also pointed out, gently, that the diagram labels a DIP switch configuration that only matters during initial install, not during troubleshooting. I'd misread the section header. The lesson isn't "don't read diagrams." It's "don't read diagrams at 9 p.m. on a Friday when you're already behind."
The turning point: the week the lounge smelled like a gym bag
August 14, 2023. I walked into the lounge and knew before I saw anything that something was wrong. That smell—warm, damp, vaguely sour. The CP70 was running. The window trim was starting to curl.
I called the vendor who'd sold it to me. He was helpful, in the way people are helpful when they don't want to issue a refund. "It's an evaporative cooler," he said. "It cools by adding moisture."
Which is true. Which is also the exact thing I should have known before I bought it. Which is the exact thing he should have asked about before he sold it to me.
That's when I stopped trusting "plug-and-play" as a category and started treating every purchase as a spec-matching exercise against the actual room, the actual climate, and the actual human beings who'd be using the thing.
The rest was math.
What I changed
I built a checklist. It's not fancy. Four questions, and I ask them before every purchase over $200.
- What does the spec sheet say, and what does it not say? The CP70's spec sheet said "cools up to 500 sq ft." It did not say "in dry climates only, adds humidity, requires ventilation." I now ask the vendor to send me the "what this doesn't do" paragraph. If they can't find one, I look for a review from someone in a similar climate or building type.
- Who is actually going to use this, and what will they assume? The cleaner assumed the Combo j9+ would handle everything. The unit was fine. The assumption wasn't. Now I print a one-page "what this does and doesn't do" sheet and walk through it with whoever will be using the device.
- What's the maintenance cadence, in actual weeks? "Monthly bin empty" on a listing means monthly. On a 12-unit property, it means weekly, or the bin overflows. I now translate every manufacturer cadence into a real operational schedule. If the answer is "more than once a week," I don't buy the unit for shared spaces.
- What's the failure mode? The CO detector had a manufacturing defect. It was caught because of a test. The Roomba 877's failure mode was a full bin. The CP70's failure mode was mildew. Every device fails. The checklist question is: "When this fails, will I know, or will it fail quietly?"
The checklist I should have had in March
If I could redo the 2023 refresh, I'd hold everything for one week and run this list first. Total time: about four hours. Total savings: the $4,200 I spent, minus the $900 or so I'd still have spent on the right equipment.
The honesty point matters more than the money. The vendor who paid attention to my humidity question—who said "actually, this won't work for that room"—was worth more than the entire evaporative cooler transaction. He didn't win the CP70 sale. He won the next three orders. That's the practical version of "transparency builds trust." Not a slogan. A working model. The vendors who told me what wouldn't work got the business. The vendors who told me what I wanted to hear got the return.
I still buy from iRobot. The 877s do their job in the smaller offices. The J9+ has held up well in the lobby now that the cleaner knows what to expect. I'm not telling you to avoid plug-and-play equipment. I'm telling you it means "plug and play, if the room matches the spec." Check the room. Then the spec. Then, only then, the price.
Simple. Not easy. Worth it.

