[{"data":1,"prerenderedAt":4},["ShallowReactive",2],{"article-body-the-descoped-still-have-homework":3},"\nIn March 2026, the EU's first Omnibus Directive on sustainability\nreporting entered into force, and with it the great descoping: CSRD\nobligations now concentrate on companies above 1,000 employees and\n€450 million turnover, releasing thousands of mid-sized firms that had\nspent two years preparing. In many of those firms, the preparation\nstopped the week the ink dried. Budget reallocated, project closed,\nrelief general.\n\nThe relief mistakes how this regulation propagates. CSRD was never\nonly a direct obligation; it is a value-chain reporting regime. The\nlarge companies still in scope must report on the impacts, risks and\nemissions of their supply chains, under assurance, on deadlines. Their\nsupply chains are, to a first approximation, the companies that just\nleft scope.\n\n## What actually changed, and what did not\n\nWhat the descoped company lost in March was the obligation to publish\nits own sustainability statement under ESRS. What it kept is\neverything operational: the customer questionnaires, the Scope 3 data\nrequests, the audit-adjacent scrutiny, because in-scope reporters'\nassurance providers increasingly test the supplier data behind the\nstatements. Survey work on CSRD implementation has consistently found\nsupplier data availability the top Scope 3 challenge, cited by around\n79% of reporters, which tells you exactly where the pressure lands\nnext. On suppliers, with sharper questionnaires and less patience.\n\nAnd the requests arrive worse than before. Under CSRD-for-everyone,\na mid-sized firm could build one ESRS-shaped dataset and serve all\ncomers. Post-descoping, the same firm faces each customer's own\nquestionnaire format, thresholds and deadlines, with no common\nstandard to anchor on. The homework survived; the syllabus was\nwithdrawn.\n\nMeanwhile the machinery most companies built for CSRD was never\nbuilt to answer ad-hoc questions anyway. Reports of CSRD-era\npractice put spreadsheet reliance among sustainability reporters\nabove 90%. Annual-cycle tooling, manually assembled, one heroic\neffort per year. Pointed at a monthly stream of heterogeneous\ncustomer requests, that machinery produces exactly what it was built\nto produce: one heroic effort, per request.\n\n## The capability, not the report\n\nThe strategic error available right now is to treat each incoming\nquestionnaire as a small reporting project. The alternative is to\nnotice that every request draws on the same underlying facts, and\nbuild the facts once:\n\n- **A data foundation, not a document.** Energy, emissions factors,\n  workforce, materials and site data maintained continuously with\n  lineage, so any questionnaire is a projection of the same audited\n  numbers rather than a fresh assembly.\n- **Supplier-of-a-supplier readiness.** The same value-chain logic\n  cascades downward: the descoped company's own suppliers will be\n  asked for their inputs. Firms that can pass structured questions\n  down and aggregate answers up become easy to keep; firms that\n  cannot become the bottleneck their customers eventually design\n  out.\n- **Consistency as the control.** When five customers ask\n  overlapping questions in five formats, the risk is five subtly\n  different answers, each individually defensible, collectively\n  contradictory, and discoverable. One source serving every format\n  is not efficiency alone; it is how you avoid contradicting\n  yourself under scrutiny.\n\n> The companies that stayed in scope get a standard, a deadline and\n> a budget line. The descoped get the same questions with none of\n> the three. The capability is how you answer cheaply anyway.\n\nThe commercial reading is simpler still. Large buyers are actively\ntiering suppliers by data reliability, and procurement processes\nincreasingly score it. For a mid-sized firm, being the supplier\nwhose numbers arrive structured, consistent and on time is a\ndifferentiation cheaper than most marketing.\n\n*The regulation moved. The questions stayed. Build for the\nquestions.*\n\nA useful benchmark: count the hours your team spent on the last\nthree customer ESG questionnaires, and how many of the answers were\nassembled fresh each time. If the second number is not near zero,\nthe foundation is missing, and building it is what the Prophesee\nESG module is for. [Serve the next request from a system](/contact).\n",1786833838254]