14 min· water· building· climate
The friction worth removing
America's water infrastructure is old, leaking and exposed, and the technology to fix most of it is already on the market. What is slow is the distance between a utility that needs something and a vendor that has it. That distance is what we are working on.
Last week I was in New York for the Impact State of Mind conference, a day on sustainable finance and impact investing held during Climate Week. The rooms were full of entrepreneurs describing what they had built, investors describing what they back, and the ecosystem partners (foundations, agencies, corporates) who sit between the two. Nearly every conversation came back to the same question, asked of every company in the room: what, measurably, is different because you exist?
Sitting there listening, I kept turning that question on my own company. AquaIntel does not build treatment plants or lay pipe. It reads documents and helps vendors sell into water utilities. It is fair to ask what that has to do with climate, and I did not have a crisp answer ready. This piece is my attempt at one, and it starts somewhere much more ordinary than a conference.
Turn on a tap anywhere in the United States and you are relying on something most people never think about until it fails: a treatment plant, a network of pipes, pumps and storage tanks, and a small public organization responsible for all of it. There are about 50,000 of those organizations for drinking water and another 16,000 for wastewater. Together they are one of the least visible and most essential systems in the country.
By the profession's own grading, they are also in poor shape. The American Society of Civil Engineers gave drinking water a C- in its 2025 report card, and wastewater a D+. Both are the same grades they received four years earlier, in a report where the country's infrastructure overall scored its best result ever.
For most of what is wrong, the fix is already on the market. Leak detection, smart metering, pressure management, advanced treatment, energy recovery, water reuse, network monitoring and industrial-grade cybersecurity all exist, and they work. The solutions exist and the vendors exist. The friction is in the two of them coming together, and that friction is what I am working on. It matters to far more people than the ones who sell into water.
What is wrong
"Aging infrastructure" has been said so often that it no longer registers, so here are the specifics.
The pipes leak. ASCE estimates a water main breaks somewhere in the country every two minutes, and that around six billion gallons of treated water are lost every day. That water was already pumped, filtered, disinfected and pressurized before it escaped. The industry calls it non-revenue water: produced and paid for, never delivered and never billed.
The bill for fixing it is large and growing. The EPA's most recent survey of drinking water needs put the requirement at $625 billion over twenty years, 32 percent higher than the survey before it. Roughly two-thirds of that is pipes.
On the wastewater side the gap shows up in the spending as well as the plan. ASCE put the annual capital need for wastewater and stormwater at around $99 billion in 2024, against about $30 billion spent.
It is exposed. I have seen utilities that still run on Windows 98. The EPA's inspections point the same way: in 2024 it reported that more than 70 percent of the drinking water systems it had inspected did not fully meet the Safe Drinking Water Act's risk and resilience requirements, and that some had critical cybersecurity gaps, such as default passwords never changed and a single login shared by all staff. These systems control the chemical dosing of the water people drink.
And it is a climate problem as well as an engineering one. Moving and treating water takes a great deal of energy. The EPA estimates that drinking water and wastewater systems account for about 2 percent of all energy use in the country and over 45 million tons of greenhouse gases a year. For a typical city, its water and wastewater plants are often the single largest energy user, at 30 to 40 percent of the municipal total.
Put those together and a leak stops looking like a maintenance issue. Every gallon lost in the ground carries the electricity that pumped it, the chemicals that treated it and the emissions that came with both. An inefficient plant burns power it did not need. A utility that cannot reuse its treated wastewater draws harder on a river or an aquifer that a hotter, drier decade is already straining. Modernizing water infrastructure is one of the most direct sustainability investments a country can make, and it mostly happens out of sight, one utility at a time.
The solutions are sitting on the shelf
I have spent most of the last seven years in this sector, first as a consultant and now building AquaIntel, and what still strikes me is how much good technology is waiting to be used.
There are acoustic sensors that hear a leak before it surfaces, meters that report consumption by the hour, software that models a pipe network and predicts which main will fail next, treatment processes that remove contaminants the older plants were never designed for, and pumps and blowers that use a fraction of the energy of the ones they replace. Much of it has been proven many times over, at utility after utility.
Adoption is still slow, slow enough that the industry has a name for the place good technology goes to die. Isle Utilities, which runs technology trials for utilities around the world, calls it the piloting valley of death: the gap between a successful pilot and a commercial contract, where proven products routinely stall because no budget, contracting route or plan to scale was ever attached to the trial.
So the useful question is why it takes so long for the technology that exists to get where it is needed. The answer is friction, and it is worth taking apart, because some of it should stay.
Where the friction comes from
Some of it is legitimate caution. When a utility adopts a new treatment process and gets it wrong, the result is a public-health event and a community that stops trusting its tap water. Being conservative about what goes into a drinking water system is the correct instinct, and nothing here is an argument against it.
Some of it is how public money moves. Utilities plan capital work years ahead, fund it through budgets, rate cases, bonds and state loan programs, and buy through public procurement with its own rules and timelines. A need raised at a board meeting might reach a budget two cycles later and a published request for proposals a year after that. By the time the RFP appears, the scope is written and the approach has usually been shaped by whoever was in the room early. I have written about that at more length.
Some of it is fragmentation. Sixty-six thousand utilities means sixty-six thousand separate buyers, each with its own board and its own appetite for risk. Isle's observation is that many insist on running their own pilot even for a technology proven elsewhere. A result at one utility transfers poorly to the next, so every sale starts close to zero.
And a large part of it, the part I think is most fixable, is that the need is hard to see.
The need is public, and it is scattered
A utility does not publish a list of its problems, because public bodies do not communicate that way. What it does instead is talk about its problems constantly, in pieces, in places nobody outside the utility reads.
The operations manager tells the board, for the third meeting running, that a pump station keeps failing. The capital improvement plan carries a line for rehabilitation in a future year. An engineering report recommends a treatment upgrade. The annual water quality report discloses a contaminant close to its limit. A consent order sets a deadline. Residents complain about pressure or discolored water, and the complaints are minuted. A bond issue names the plant.
Every one of those is in the public domain. None of them, on its own, says this utility is going to need something in the next two years. Together they say it clearly to anyone who has read all of them, in order, and noticed they are the same story.
That is very hard for a person to do for one utility, and impossible for sixty-six thousand. The documents are spread across tens of thousands of websites, in every format from searchable PDF to scanned image, with no common structure and no index.
The picture is often not assembled on the inside either. About 81 percent of community water systems serve fewer than 3,300 people. Many run with a handful of staff who are fully occupied keeping the water flowing. They know their system intimately, but they do not always have the time or the specialists to turn what they know into a clear, prioritized statement of need, and they certainly do not have time to evaluate every vendor who might help.
What the friction costs
On the vendor side, the cost shows up first as time. The complaint I hear from almost every vendor selling into this sector is the length of the sales cycle: anywhere from six to twenty-four months from first conversation to signed contract, depending on the product or service. I believe that is too long, and most of it is not needed. Some of it is the public process doing its job. A good deal of it is effort spent in the wrong places.
Most people in this industry will tell you it is a relationship game, and they are right that relationships matter. But a relationship is one tool in the toolkit. Without a way to see which utilities have a relevant need, a sales team spends its relationship time by habit: it works from population tiers and old contacts, attends the conferences, answers the RFPs that appear, and has a great many conversations with utilities that were never going to buy what it sells. The data does not replace the relationship. It tells you which ones to build, including with utilities you would never have thought to approach.
On the utility side, the cost is attention. Every one of those poorly targeted conversations lands on a small team with no time to spare. A utility manager who has sat through enough pitches for things they do not need learns to stop taking the meetings, eventually including the one that would have helped.
And the cost that matters most falls on neither of them. It falls on the pace of modernization itself. When it takes years for the right product to find the right utility, the leak runs longer, the old pump burns more power, the vulnerable control system stays online, and the plant that could have been reusing its water keeps discharging it. Multiply a few years of delay across tens of thousands of utilities and it becomes a meaningful share of the gap between the infrastructure the country has and the one it needs.
What we are building
AquaIntel exists to take as much of the avoidable friction out of that process as possible. The caution and the public process are there for good reasons, and we leave them alone. What we work on is the part where the need is public and nobody can see it.
We read the public record at the scale of the whole country (board minutes and agenda packets, capital plans, budgets, engineering reports, compliance data, funding records) and assemble the scattered pieces into one picture of what is going on at each utility: what it owns, what condition it is in, what it has committed to spend and when, what it is being required to do, and who decides.
That picture is useful in two steps.
The first step is knowing who to talk to. Across the whole market, which utilities genuinely have a need that a particular vendor's product addresses, and why, with the documents that show it. That is screening, and it removes the single biggest source of wasted effort in this market at the start. A vendor that talks to twenty of the right utilities instead of two hundred of the wrong ones saves its own time, and saves the time of the hundred and eighty utilities it did not need to call.
The second step is arriving well informed. For each of those utilities, a full account picture: the history of the problem, the budget behind it, the people involved in the decision, and what has changed since anyone last looked. What that buys is a better first conversation, one where the vendor already understands the utility's situation and the utility does not have to spend the first three meetings explaining it.
That is better for both sides, and the utility's side is the one I care about more. A utility that meets a vendor who has done the reading is being treated as what it is: a public body with specific, documented needs.
Why this is a sustainability argument
The contribution, if it works, is indirect. We do not replace pipes or treat water. What we can do is shorten the distance between a utility expressing a need and the right solution arriving to meet it. Every month taken off that distance is a month sooner that a leak is found, a pump is replaced with an efficient one, a plant starts recovering energy or reusing water, or a control system is locked down.
That is a real lever, because the need is so large and the solutions already exist. When the constraint on progress is adoption rather than invention, the most useful thing you can do is make adoption faster. The money is being spent either way, hundreds of billions of it over the next twenty years. Whether it is spent on the best available technology, at the right utilities, in the right order, depends a great deal on whether the people buying and the people selling can find each other early enough to shape the decision.
What we cannot claim yet
We cannot measure any of this yet. The workflows exist, customers use them, and the early signs are the ones you would hope for: conversations that start earlier, with fewer wrong turns. What we do not have is the evidence the argument above rests on, which is that a utility modernized sooner, or better, because a vendor found it through us.
That is measurable, and there is an obvious place to start: the sales cycle. If the six to twenty-four months vendors describe today does not come down for the vendors who work this way, we are not removing the friction we say we are. It is the number we will track, and the one to hold us to.
It is not the whole story, though. The record that tells us a need is forming also tells us, eventually, what was bought, when, and from whom. Holding both ends of that story, the first signal and the final award, is how we will find out whether we are shortening the gap or only making the selling more efficient. Only the first of those is the reason I am doing this.
So my answer to the question that kept coming up in New York is that the impact is not yet measured, but it is measurable, and we are building the means to measure it.
Where to start
If you sell into utilities, one exercise is worth doing with your team this week. Take your last three wins and put two dates next to each: the day you first heard about the need, and the date of the first public document that described it (a line in the minutes, a capital plan, an engineering report). The gap between those two dates is the part of your sales cycle this essay is about.
If you want to see what that looks like for your own territory, reach out. I am happy to walk you through it.
Related
- The problem statement I couldn't unsee: how I came to this problem.
- Nobody reads the minutes: what the scattered record looks like when someone does read it.
- Fifty thousand is not a market: step one, in detail.