A company can meet every water target it set, publish a defensible volume of water replenished, and still operate in a basin that produces less water than it did the year the program started. Nothing in that sentence is greenwashing. It is a scale problem, and it is the quietest failure mode in corporate water work.
Corporate water stewardship is the best-organized part of corporate nature management. It has methods, validated targets, a volumetric unit, and a claims standard — more machinery than biodiversity or land has. What it does not have, and what no accounting framework can have, is a budget line that owns the physical condition of the ground where your water is made — usually someone else's, if anyone's.
what corporate water stewardship actually is
Corporate water stewardship is the practice of managing a company's water use, water quality impacts, and water-related risk beyond its own fence line — at the site and in the shared basin — on the premise that water risk is shared and cannot be solved inside a property boundary.
In practice the program runs five moves:
- Measure the footprint — withdrawals, consumption, and discharge across operations and, harder, the supply chain.
- Prioritize by basin — which sites and sourcing regions sit in water-stressed catchments.
- Set targets — context-based rather than a flat corporate percentage, ideally validated by someone outside the company.
- Act inside the fence, then beyond it — efficiency and reuse first, then projects in the basin: managed recharge, irrigation efficiency, leak repair, riparian restoration, WASH access.
- Account and claim — quantify the benefit and report it against the target.
That is real work, and mostly good work. It is also why the vocabulary gets slippery: every one of the five moves produces a number, and none of the five numbers is the basin.
the three frameworks your water lead is already holding
SBTN — the target. The Science Based Targets Network's current freshwater method covers water quantity and nutrient pollution from nitrogen and phosphorus. Groundwater withdrawal already counts on the quantity side, with explicit groundwater thresholds available only where a local model exists; Version 2 adds a dedicated groundwater target and extends the method to pesticide-related impacts. Companies including Danone, General Mills, H&M Group, Arla Foods, Decathlon, and Metso began piloting the updated guidance in June 2026, ahead of a launch later in 2026. The practical detail for anyone stalling: SBTN says validation services for current method versions stay open for six months after V2, so waiting for the new version is not a reason to hold the program.
NPWI — the claim. Net positive water impact, pioneered by the Water Resilience Coalition under the UN Global Compact's CEO Water Mandate, means a company's contributions exceed its impacts on water stress in the same water-stressed basin, across three dimensions: availability, quality, and accessibility. It is set at enterprise level and implemented at site and basin level. The coalition's 2030 ambition is positive water impact in more than 100 water-stressed basins supporting over 3 billion people, plus resilient WASH for 300 million. Note the definition's own words — contributions exceeding impacts in the same region. The framework is the one insisting that where matters.
VWBA — the unit. Volumetric Water Benefit Accounting, published by WRI in 2019 and updated as VWBA 2.0 on September 17, 2025 by WRI, LimnoTech, Bluerisk, and Bonneville Environmental Foundation, is the six-step method for quantifying the volumetric water benefit of a stewardship project — in liters, cubic meters, or gallons — and communicating the resulting claim. VWBA is what turns "we replenished more water than we used" from a press line into something a third party can check.
These three are cousins, not opponents. Ensurance does not set targets, does not validate a claim, and does not issue a volumetric unit. Read what follows as a scale note from the other side of the same problem.
where the program and the place part ways
Start with the map, because the map is where the mismatch is easiest to see and hardest to notice.
The 100 Priority Basins list is drawn at HydroBASINS Level 4. The implementation guidance is candid about what that means: Level 4 is "very coarse" and "covers areas of major basins (e.g., Mississippi or Ganges)," offered as "a recommended point of initial orientation," with companies "encouraged to define their NPWI actions at finer HydroBASIN resolutions." The framework is not pretending the polygon is the place. But a program built at that resolution reports at that resolution, and a board can read a basin-count as though it were a hydrograph.
Now the physical side. Water that shows up in a river in August was, in snowmelt-driven mountain systems, infiltrated somewhere in April or May. Infiltration is the process the whole program is ultimately serving: rain and snowmelt entering soil rather than running off it, measured in inches or millimeters per hour on a specific hillslope with a specific soil, a specific compaction history, and a specific amount of living root in it. Compact that hillslope with hooves or equipment and the same storm leaves as a flash instead of a recharge. Incise the stream below it and the meadow that used to hold water into July drains by June. Burn the canopy above it and the snow arrives in one pulse and then not at all.
Add a second process, because dependents feel this one directly: riparian shade. A reach with a shaded, connected channel delivers cooler water later in the season. The same volume through a bare, widened channel arrives warmer, earlier, and carrying sediment your treatment plant has to remove. In a volumetric account those are the same liters. In a pipe, an intake, or a cooling loop, they are not the same water.
That is the whole argument, and it is not one the accounting people would dispute. The original VWBA working paper says it directly: estimating volumetric benefits alone "cannot provide assurance that shared water challenges are reduced and social, economic, and environmental benefits are provided." The people who built the unit told you first that the unit is not the outcome.
the program and the place, side by side
| the stewardship program | the basin | |
|---|---|---|
| what it produces | a validated target, a volumetric claim, a disclosure line | infiltration, storage, and release timing |
| unit | liters, cubic meters, a basin count | condition — infiltration rate, soil carbon, canopy, floodplain connection, channel shade |
| resolution it is drawn at | enterprise, then site, then a coarse basin polygon | a hillslope, a reach, the meadow above your intake |
| time frame | a reporting year, a target date | decades |
| who is obliged | the sustainability function, with an annual budget | no single balance sheet — a watershed council, utility, agency, or tribal program where one exists, and no one where none does |
| what success looks like | contributions exceed impacts, in the same basin, on paper | the stream still running in August |
| how it fails | benefits booked in a basin your operations do not drink from | every number checks out on a shrinking total |
The two columns are not rivals, and a good water lead needs both. The mistake is reading a completed left column as a delivered right column.
Sit with the obligation row, because it is easy to misread. Where a body already exists for the source, condition gets funded and has been for decades: Denver Water's Forests to Faucets cost-shares forest treatment in its own supply watersheds, New York City's Catskill and Delaware program has paid upstream landowners and communities rather than build a filtration plant, and irrigation districts, state agencies, tribal water programs, and watershed councils carry versions of the same work. None of that is broken, and none of it needs replacing. The gap is coverage and access: most reaches have no such body, and where one exists it rarely has an open line for a downstream dependent who wants to pay into it.
the wrong-basin problem, in one sentence
A water credit or volumetric benefit bought in a basin your operations do not draw from is not a fake — it is real water in the wrong place, and it does nothing for the intake that is actually at risk. What a water credit actually is takes the instrument apart, who buys a water credit names the payors already covering shortage, and water accounting is not a restored cycle covers why a ledger cannot close the gap. This post does not re-argue any of that.
you already pay for this basin — just later, and to someone else
Each of the three groups reading this already has a nature line item. It is booked under other names.
- Corporations: trucked water during a curtailment, a production line slowed to a permit condition, a higher treatment bill after a burn upstream, a sourcing region that quietly went from one crop to another.
- Utilities: post-fire turbidity events, chemical and energy cost per million gallons, non-revenue water, and the capital project that gets approved because the source kept degrading.
- Data centers: the cooling water a permit ties to a stressed aquifer, the local moratorium or siting fight that follows, and the energy penalty of switching to air cooling to avoid both. Data centers drink water covers that draw in detail.
You already have a nature line item. You just pay it after the living system fails.
The interesting question is not whether to spend. It is whether one of those existing line items can be pointed at the ground that shrinks the next bill. Source protection is usually the cheapest water a utility will never build runs that math on the utility side, and who pays to create supply covers who funds new water when the source is not enough.
what funding the place looks like
The industry name for a beneficiary paying a steward to keep a flow coming is payment for ecosystem services, and it long predates us — what happens when one company is the only buyer is the cautionary version of that arrangement.
Ensurance is one way to write that payment against a named place rather than a unit. In one gloss: an agent is an account standing for a specific place, people, or purpose. A certificate — specific ensurance — is issued one-to-one against a single agent and funds that place's present condition, with proceeds routed to whoever does the work on the ground. Coins are the indirect route, where trading activity funds protection across the protocol. That is the whole vocabulary lesson.
What a certificate is not: a water right, an allocation, a priority date, a delivery obligation, a volumetric benefit unit, or a neutrality claim. It carries no acre-feet. It is a hold on the condition of a place — money committed to a place's present condition, not to a unit it produces — funded now rather than after the loss. If your program needs the volumetric number for a project you fund that way, VWBA 2.0 is still how you would estimate it — these are different layers, not competing products.
Underneath the funding sits valuation: the condition and service-flow accounting that lets a finance committee see a landscape as something other than a cost. We run that engine because a price has to be legible before anyone can approve it. It is a bridge for capital, never a statement of what the basin is worth. A basin is not a program milestone and not a portfolio position. Water shows up in August or it does not, and the reasons are physical.
where we actually are
Live and small: 191 coins, 26 certificates, and roughly 2,080 agents — many of them accounts for places nobody has funded yet. We do not run a water market, hold water rights, issue acre-foot credits, or sell software seats. A first engagement is a conversation about one named place and one payable, which may become a valuation, a certificate on a named agent, or a scoped services engagement. Anyone describing a nature finance product as finished is describing a roadmap.
pick the basin your operations actually drink from
Not all 100. Not the enterprise footprint. One basin, and inside it, the reach or the upland that decides your intake.
- Corporations: what this looks like for a company with a nature dependency
- Utilities: source condition as a supply decision
- Data centers: the draw, the basin, and the license to operate
frequently asked questions
what is corporate water stewardship?
Corporate water stewardship is the practice of managing a company's water use, water quality impacts, and water-related risk beyond its own operations — at the site and in the shared basin. It typically runs five steps: measure the water footprint, prioritize water-stressed basins, set context-based targets (often through SBTN), act inside and then beyond the fence line, and account for the benefit (often through VWBA). It is a management program. It is not, by itself, a change in how much water a landscape produces.
does a water credit restore a basin?
No. A water credit or volumetric water benefit records a volume attributable to a stewardship activity; it does not by itself change the condition of the land that captures, stores, and releases water. A benefit generated in a different basin does nothing for the catchment your operations depend on, which is why net positive water impact is defined as contributions exceeding impacts in the same water-stressed basin. Restoring a basin means changing physical condition — infiltration, floodplain connection, canopy, channel shade — and holding that change for decades.
what is net positive water impact?
Net positive water impact (NPWI) is a leadership ambition, pioneered by the Water Resilience Coalition of the UN Global Compact's CEO Water Mandate, in which a company's measurable contributions exceed its impacts on water stress in the same water-stressed basin across three dimensions: availability, quality, and accessibility. It is set at enterprise level and implemented at site and basin level. The coalition's 2030 ambition covers more than 100 water-stressed basins and over 3 billion people.