I’d rather buy one robot vacuum with a self-emptying dock than three budget-priced units that need constant attention. That’s not a tech opinion; it’s a spreadsheet opinion. I’m the procurement manager at a 40-person commercial cleaning company. I’ve managed our equipment and supplies budget—about $180,000 annually—for the past seven years, negotiated with more than 20 vendors, and documented every order in our cost-tracking system. So when I say prevention beats repair, I’m not reciting a slogan. I’m describing what the invoices show.
What the iRobot Roomba 985 Review Didn’t Tell Me
Before our pilot purchase, I searched for an irobot roomba 985 review and found plenty of spec comparisons. What I didn’t find were lifecycle numbers. So I started tracking our own.
We ordered two units for the pilot—or rather, we ordered one for our main client office, then a second for the satellite site after the first week looked good. The first ran five nights a week for about 13 months before it needed a new battery. The second is still running at maybe 15 months now; I’d have to check the maintenance log. That’s not a big sample, but it’s the data I have.
What impressed me wasn’t the cleaning map. It was the app reporting. We could see which zones had recurring misses and adjust the schedule before a client noticed. That’s prevention, not repair. It’s the difference between spending five minutes checking a dashboard and spending a morning rescheduling a cleaning crew.
One honest caveat: I don’t have hard data on how the 985 compares to newer units for object detection or carpet type coverage. I can only speak to our floor surfaces—carpet, tile, and vinyl—and to our night-cleaning schedule.
Why I Stopped Shopping by Sticker Price
In Q2 2024, I reviewed 18 months of service tickets across 11 robots in our fleet. The pattern was uncomfortable: the units with the lowest upfront number showed up more often in the repair pile. I wish I had tracked downtime from day one. What I can say anecdotally is that the gap was obvious.
I almost approved a $350 online distributor quote in 2023. The specs looked fine. Then I built a simple cost calculator and added in filters, likely battery replacement, and the labor hour our tech would lose un-jamming a brush. The total landed within $80 of an iRobot model with a self-emptying dock. So glad I caught that before signing.
My formula isn’t fancy: purchase price + 24 months of consumables + expected monthly labor hours × loaded labor rate. It’s not perfect. But it catches the cases where a $350 unit quietly becomes a $900 unit.
The self-emptying dock is the purchase decision I’m most confident about. It sounds like a minor convenience, but it removes a human failure point. If the robot doesn’t need to be emptied every night, nobody has to remember to do it. Same with smart mapping: when a client’s office layout changes, we update the map remotely instead of sending someone to push buttons on the robot.
Five Minutes of Checking Beats Five Days of Fixing
Our maintenance checklist now covers every piece of equipment, not just robots. The 12-point inspection I created after an early mistake has saved us thousands in avoidable rework. Five minutes of verification beats five days of correction—that’s not a motivational poster. It’s from a vendor spreadsheet I still use.
It’s strange how the same lesson shows up in unrelated searches. One of our techs needed the Yale smart lock user manual at a client site because the door code was glitching and nobody had documented the reset procedure. He spent 10 minutes finding the manual and one minute fixing the lock. Another week, we needed a handheld vacuum sealer nearby to re-seal damaged supply bags before they gathered dust. And one of our staff members looked up how to treat a curling iron burn on hand after a small incident in the break room. None of these are robot vacuum topics, but they all point to the same thing: do a little prevention now or a lot of correction later.
What I Don’t Know—and Why I’m Okay Saying It
I don’t have hard data on industry-wide robot vacuum failure rates. I can only speak from our fleet. I also don’t have a lab, so I can’t guarantee the 985 works on every floor type. It works on our carpet, tile, and vinyl, and I’ll leave it at that.
Pricing and model lineups change quickly. This was accurate as of mid-2025; iRobot updates its lineup, competitors adjust their features, and availability shifts. Verify current specs before you present a purchase to your finance team.
One more thing: per Federal Trade Commission (FTC) guidance (ftc.gov/business-guidance/advertising-marketing), advertising claims have to be truthful and substantiated. But “substantiated” doesn’t mean “true in your building.” That’s why I treat every vendor promise as a hypothesis to test, not a fact to repeat.
But a Basic Model Is Fine for a Small Office, Right?
Maybe. I can hear that objection. For a single reception area with light traffic, a basic robot might be enough. But in our experience, a small space is exactly where a stuck robot matters most. There’s no night team to catch it, and there’s often no backup plan.
The counterintuitive part: the bigger our client site, the less risky a basic robot becomes, because someone is around to fix it. The smaller the site, the more reliable the equipment needs to be. That flipped my purchasing logic. In a small office, I want the most dependable unit, not the most affordable one.
What I’d Buy Next
If I had to roll out another small-office fleet today, I’d still put the Roomba 985 in as the workhorse, then add a dedicated mop for hard floors. The 985 isn’t perfect. I can’t speak to every floor type, and I won’t pretend it’s the best at every cleaning task. But it’s predictable, and predictability lowers our labor cost.
I’d also look at iRobot’s newer vacuum-mop combos for the sites that need one device instead of two. But we don’t replace equipment that isn’t broken. The 985s are still running, so they stay.
Stop Buying Repair Cycles
So here’s my view after seven years of purchase orders: don’t ask “what’s the most affordable robot vacuum that looks good in a search list?” Ask “what prevents the next unplanned visit?” If a purchase can’t be tied to a future problem it prevents, you’re not buying equipment. You’re buying a repair cycle.
I know that sounds like something a finance person would frame and hang on the wall. (I’m a finance person, so fine.) But it applies beyond robot vacuums. A lock manual you document today, a sealer you have on hand, a first-aid checklist in the break room—these are all ways of saying the same thing: prevention is the cheaper option.
Discuss this topic with Irobot