Watershed
Each of these is a real decision Watershed is working through, written so you can make the call yourself — incomplete information, someone who pushes back, no answer key.
Watershed provides an enterprise SaaS platform that helps large corporations measure, report, and reduce greenhouse gas emissions across scopes 1, 2, and 3. Customers use it to drive ESG insights, report on disclosures, and decarbonize their businesses and supply chains. It serves Fortune-scale customers such as Walmart, Airbnb, Stripe, and DoorDash, and its customers include more than 20% of the Fortune 500. Its Financial Institutions team customizes the platform for Asset Managers, Banks, Portfolio Companies, and other financial institutions. For them the platform models finance-specific entities such as portfolios, assets, and holdings, calculates data quality metrics like PCAF scores, and supports customer reporting on financial regulations such as SFDR. Watershed also offers AI Products, including Product Footprints and an AI Platform for sustainability.
Your first one is free. No card.
13 live right now.
The rest
Map shared metrics across filings without treating them as identical
Authors are rebuilding metric maps outside the reporting application because one emissions value carries different context across filings. You have to weigh faster reuse against making every framework-specific exception visible.
Hold disputed claims before the filing package locks
Reviewers agree on the emissions numbers but disagree on what drafted language promises about them. You have to weigh a faster filing path against a review flow that preserves source support and accountability.
Clarify reusable answers before authors rewrite them by hand
Authors are abandoning reusable metrics because the picker hides the context they need before insertion. You must weigh faster placement against enough provenance for authors to trust the answer they reuse.
Compare portfolio changes without rebuilding the analysis in Excel
Reporting managers are exporting portfolio data because proposed holding changes and PCAF scores cannot be compared together. You have to weigh a useful first scenario workspace against results users can trace back to the underlying data.
Rank accounts by renewal risk with checkable evidence before reviews
Strategic accounts are reaching executive reviews with unresolved data decisions and missed commitments discovered too late. You must weigh earlier risk ranking against the evidence and human review needed for customer-facing action.
Build an evidence trail for eight pilots instead of bespoke work
Pilot customers can file emissions numbers without a reusable trail explaining how the data was transformed. You have to weigh a guided programme’s value against customer capacity and a limited advisory review window.
Decide what a filed record must preserve
In a filing rehearsal, 16 of 100 statements could not be rebuilt as filed once ordinary edits landed. You have to weigh audit defensibility against letting customers correct their work, before the mandated cycle.
Pick the method you can defend at filing
A rehearsal shows 23% of the financed-emissions figure rests on modelled borrower data. The audit committee approves only one methodology, so you have to choose the evidence standard you can defend at filing.
Keep the drafts inside what the sources support
Red-team checks found 6 of 180 polished statements merged assumptions from documents that conflict, with no signal to the reader. You have to settle on a model and a review design before reporting season.
Settle the terms before the 18-day deadline
A global buyer wants one binding commitment on data access, service levels, pricing, and liability, with a decision in 18 days. Whatever you concede becomes the template for their global program.
Cut over boundary changes without losing a facility
A rehearsal found eight boundary changes where facility membership in the approved register does not match what the calculations use. You have to weigh how far to verify before new reporting runs start.
Decide if the cutover can go ahead
Two of three business units reconciled their sampled totals, but one has unresolved mapping exceptions and reviewers cannot agree on tolerance. You have to decide whether to cut over or risk parallel reporting during the close.