Fifty thousand is not a market
There are around fifty thousand drinking water utilities in America and sixteen thousand more on the wastewater side, and no vendor sells to all of them. Screening is the workflow that turns a number that large into a list a person can actually work — and the honest answer is far shorter than anyone expects.
There are roughly fifty thousand community drinking water systems in the United States and sixteen thousand wastewater utilities alongside them. Every vendor selling into this market knows that number, quotes it in their board deck as the size of the opportunity, and then does business with a few dozen accounts.
The gap between those two facts is where most go-to-market effort in this sector is quietly wasted.
The job, as it exists today
Somebody has to decide which utilities are worth anything to you.
Not which are large — that is easy and mostly wrong. Which ones have the problem you solve, at a moment when they are able to do something about it, in a place you can serve, at a size that justifies the trip.
Almost nobody does this deliberately. Territories get drawn from history, from where a rep happens to live, or from a list somebody bought. The result is sales teams working accounts because they have always worked them, in a market where the interesting ones are almost by definition the ones nobody has worked yet.
Why it is harder here than it sounds
The market's shape is hostile to filtering. I wrote about this before I started the company: the American water sector is a long tail with almost everything at the thin end. A hundred and thirty-eight utilities serve ninety-nine million people. More than forty-three thousand serve twenty-four million between them. Sort by population and you get the same two hundred names every vendor already knows. The rest of the market — most of it, by count — is invisible to any filter built on size.
And no two vendors want the same cut. The obvious filters are the ones everybody already has: population, state, treatment type. The filters that actually separate a good account from a bad one are specific to what you sell. Miles of pipeline. Meter type and age. Whether they have said anything, ever, about a contaminant you treat. Whether they own their plant or contract its operation. Whether they have money.
Those are not fields in a database. They sit in documents, which is why the monitoring workflow comes first in this series — screening on anything interesting is downstream of having read everything.
What the workflow does
You define what good looks like. The criteria are the customer's — the same principle as the fit-check, and for the same reason. We are not in a position to know which utilities are right for your business. You are, and usually in some detail; what you have lacked is a way to apply that knowledge across sixty-six thousand organizations rather than the forty you can hold in your head.
We run it against real depth, and I should be precise about how much. We hold in-depth data on around six and a half thousand utilities. That is a tenth of the count and roughly eighty per cent of the American population served.
Depth follows population, because the larger a utility is the more it publishes and the more consistently it does so. So a screen runs against substantial material for the utilities that account for most of the market, and against less as the systems get smaller. The true long tail — the forty thousand small systems — is thinner for us as it is for everyone, and it is the part of the record we are still extending.
Values
| Tier | Utilities, by count | Population served |
|---|---|---|
| Held in depth | 6,500 | 80% |
| Not yet | 59,500 | 20% |
Different lenses, not one filter. The same market screened for a metering company and for a treatment chemicals company produces two lists with almost no overlap. So the workflow is not one saved search; it is a set of them, defined per product line, per campaign, per territory.
Territory is one of those lenses rather than a filter bolted on afterwards. You define your sales regions however your business actually divides them — which is rarely by state line — and the results come back cut that way.
And the output narrows as you go. The results are a dashboard you work in. What survives that can come out as a CSV, because plenty of real work still happens in a spreadsheet and pretending otherwise helps nobody. And for the shortlist — the accounts you have decided are worth the attention — you generate the full profile, which is where the reading workflow takes over.
That progression is the argument of this article expressed as a product: many, then fewer, then a small number understood properly.
And when we do not have something, we say so. If you ask for a criterion we do not hold — through the search bar, in plain language — the answer is that we do not have it, not a quiet approximation dressed up as a result. Usually alongside a suggestion: here is a related signal that tends to move with the one you wanted, if it is any use to you.
That is the third time this principle has appeared in this series, after candidate-versus-verified contacts and the visible filtered-out queue. Not a coincidence, and by now something closer to a house rule: a system that hides the edges of its own knowledge is a system nobody can calibrate against.
One company, two answers
The most useful thing anyone said to me about screening came from someone describing two businesses inside the same parent company.
One of them sold equipment with a large installed base, most of it approaching renewal. That business understood its pipeline well: it knew where its machines were and roughly when each would come up, and screening for it meant confirming and timing something already half-known.
The other sold a newer product into a market it had barely entered. It had no installed base to reason from, no renewal cycle, and very little idea which utilities should even be on the list.
Same company, same market, same fifty thousand utilities — and two completely different jobs. The first needed a calendar; that is the installed-base workflow. The second needed a map of somewhere it had never been.
Any screening tool that assumes every user is in one of those positions will be useless to the other, which is most of why generic prospecting products do not work in this sector.
The honest answer is a short list
The natural output of a screening system is a big number. Five hundred qualified accounts sounds like value delivered; it justifies the subscription and it looks impressive in a review. It is also, in this market, mostly a way of feeling productive.
A vendor with a couple of dozen field people cannot meaningfully pursue five hundred utilities. What they can do is know twenty of them properly — what those utilities are struggling with, who decides, what they have already tried, when their contracts come up — and be genuinely useful to those twenty. That is the argument I made about not spraying and praying, arriving at the same place from the other end.
So the screen that matters is not the one that returns five hundred names. It is the one that gets you from sixty-six thousand to a few hundred worth considering, and then tells you which twenty to start with, and why.
The rest of this series is what you do with those twenty.
What it cannot do
It will not invent an answer. Ask for something we do not hold and you are told so. The failure mode being avoided there is specific and worth naming, because it is the default behavior of almost every system of this kind: quietly returning a list, because a list is what was asked for, and letting you assume it means what you wanted it to mean. A wrong list is far more expensive than no list. It gets acted on.
What you get instead is a view — the adjacent criterion, the nearest available proxy, the signal that tends to move with the one you asked for — and then you decide whether that is any use.
And a list is not a plan. Screening produces attention, not action. Every other workflow in this series exists because a list of good accounts, handed to a sales team with nothing else attached, ends up exactly where the bid feed ended up: in a folder nobody opens.
Next
Screening ends with a shortlist, and a shortlist is only a set of names until somebody understands what is actually happening inside each one. The next piece is the deep account profile — the thing this company started with, before there was a platform or a workflow or a plan: everything worth knowing about a single utility, assembled in one document.