The Corporate Sustainability Reporting Directive requires Scope 3 disclosure under ESRS E1 (Climate Change). The standard references all 15 GHG Protocol Scope 3 categories. What that actually means in practice, for a company that has never filed a CSRD report before, is considerably more nuanced than the regulation text suggests.
We've worked through ESRS E1 disclosure requirements with several growing businesses preparing their first filings under the CSRD phased schedule. The process is harder than the standard lets on, and the hardest part isn't the math. It's knowing which categories to report, at what quality level, and how to document the boundary decisions your auditor will scrutinize.
What ESRS E1 Actually Requires on Scope 3
ESRS E1-6 is the disclosure requirement for gross Scope 1, 2, and 3 greenhouse gas emissions. The standard does not let you pick and choose which Scope 3 categories to report. Paragraph 44 of ESRS E1 requires disclosure across all relevant Scope 3 categories using the GHG Protocol Corporate Value Chain (Scope 3) Standard as the measurement methodology.
The phrase "all relevant categories" carries legal and audit weight. You are required to assess all 15 categories and either report emissions for each relevant one or document, with supporting reasoning, why a category is not material to your value chain. A blanket "not applicable" without documented assessment will not satisfy an assurance engagement under ISAE 3000 or IAASB's recently finalized sustainability assurance standards.
This is where most first-time CSRD reporters underestimate the work. The assessment of relevance is itself a documented deliverable, not a checkbox.
The 15 Categories: Where the Data Challenge Actually Lives
The GHG Protocol organizes Scope 3 into 15 categories split between upstream (Categories 1-8) and downstream (Categories 9-15). The upstream categories cover purchased goods and services, capital goods, fuel and energy activities not in Scope 1/2, upstream transportation and distribution, waste generated in operations, business travel, employee commuting, and upstream leased assets. The downstream categories cover transportation and distribution of products sold, processing of sold products, use of sold products, end-of-life treatment of sold products, downstream leased assets, franchises, and investments.
In our experience working with manufacturing and B2B services companies, the data challenge concentrates in three areas:
- Category 1 (Purchased goods and services): Typically the largest category by tCO2e volume, and the hardest to measure with supplier-specific data. Most companies start with spend-based emissions factors, which introduces 30-60% uncertainty depending on your supplier mix and the factor library used.
- Category 11 (Use of sold products): For any company selling energy-consuming goods or materials that get processed downstream, this category requires modeling end-use scenarios. The data is inherently synthetic unless you have customer usage data.
- Category 4 (Upstream transportation and distribution): Requires knowing freight distances, modal mix, and load factors. Most mid-size companies don't have this sitting in a single system.
Year One: What You Can Reasonably Omit (and How to Document It)
We are not saying you should skip categories. What we are saying is that ESRS E1 explicitly accommodates phased improvement through its materiality and significance thresholds, and your materiality assessment is the legitimate mechanism for scoping year-one disclosure boundaries.
Under ESRS 1 General Requirements (paragraphs 22-30), materiality for sustainability matters is assessed using both impact materiality and financial materiality lenses. A Scope 3 category can be assessed as not material if you can demonstrate the emissions volume is likely immaterial relative to your total footprint and the associated impacts are not significant for your value chain. That assessment must be documented in your ESRS E1 disclosures alongside the basis for the conclusion.
For a B2B software company preparing its first CSRD filing, Categories 10 (Processing of sold products), 12 (End-of-life treatment), and 13 (Downstream leased assets) may genuinely be not applicable or immaterial. The documentation burden for those is relatively low: describe the category, explain why your business model means it contributes negligibly, and note the basis for the estimate or exclusion.
Where year-one omissions become risky is when a company excludes a category that is obviously significant (e.g., a logistics company excluding Category 4, or a manufacturer excluding Category 1) on the grounds that data collection is difficult. Difficulty of data collection is not a materiality argument. Auditors will push back on this.
A Practical Data Collection Sequence
Here is the sequence we recommend for companies beginning CSRD Scope 3 data collection, based on a realistic data maturity timeline.
Start with spend-based estimates for Categories 1-8. Use EXIOBASE or the EPA USEEIO supply chain factors (or DEFRA's supply chain factors for UK-heavy purchasing patterns) applied to your procurement spend by category. This gives you an order-of-magnitude view within 6-8 weeks and tells you which categories are likely to be material before you invest in primary supplier data collection.
Layer in activity-based data for high-materiality categories. If Category 1 is 70% of your Scope 3 (a common finding for manufacturers), that is where you focus supplier-specific data collection, even if it only covers your top 10 suppliers by spend. The GHG Protocol allows hybrid methods: use supplier-specific data where available, spend-based factors for the long tail.
Document your factor choices and their vintage. ESRS E1 requires you to disclose the calculation approach, the emission factors applied, and their source. A factor citation like "DEFRA 2024 conversion factors" is acceptable. A factor cited as "internal estimate" without a source is not.
A Scenario from Year-One Reporting
Consider a growing industrial equipment manufacturer preparing its first CSRD filing under the large-company schedule. Their total Scope 1 and 2 footprint runs about 18,000 tCO2e annually. Initial spend-based Scope 3 analysis puts Category 1 alone at approximately 140,000 tCO2e, with steel procurement accounting for roughly 60% of that figure. Categories 11 (use of sold products) and 12 (end-of-life) are genuinely negligible because the equipment operates on compressed air, not electricity, and is typically refurbished rather than scrapped.
Their year-one disclosure covers all 15 categories: Categories 1-7 with quantified estimates (3 using primary supplier data, 4 using spend-based factors), Category 11 with a quantified estimate using representative operating scenario modeling, and Categories 12-15 with documented not-material assessments explaining the business model basis. The report includes a data quality table showing the estimation method and factor source for every category quantified. That package satisfies limited assurance requirements under ISAE 3000.
The key decision they made: prioritize getting steel suppliers on a data collection program in year two, because Category 1 steel emissions are the material number that will get scrutinized in every subsequent filing.
What the Auditor Actually Checks
Limited assurance under CSRD does not mean a light-touch review. For Scope 3 specifically, an assurance provider will check that all 15 categories were assessed (not just the ones you reported), that the methodology is consistent with GHG Protocol Scope 3 Standard, that factor choices are documented with source and vintage, and that double counting has been considered at category boundaries (particularly between Categories 1 and 2, and between Categories 4 and 9).
The chain-of-custody requirement is particularly important: for every tCO2e figure you disclose, the auditor wants to be able to trace the number back to a calculation. That calculation must reference an activity datum (spend, distance, quantity) and an emissions factor. Both must be recorded and retrievable.
This is why building a ledger that records the full calculation lineage, not just the final figures, matters for CSRD compliance. A spreadsheet that shows totals but not the individual rows backing each category total will create friction in any assurance engagement.
How We Handle This in Emitpulse
When a company connects their procurement data to Emitpulse, we run the Category 1 spend-based estimate first. The output is a category-level breakdown showing estimated tCO2e per Scope 3 category, the spend amount applied, the factor source and vintage, and the methodology flag (spend-based vs. activity-based vs. supplier-specific). That breakdown becomes the foundation of the ESRS E1 category table in the disclosure report draft.
For categories where we do not have data to quantify, we generate a not-assessed or not-material template note in the disclosure draft that the sustainability manager can review and modify before the auditor sees it. The audit trail records which categories were quantified and which were excluded, and why.
We're not claiming this fully solves the hard work of Scope 3. Getting Category 1 to supplier-specific data quality requires your suppliers to actually respond to emissions surveys, and that is an engagement problem, not a software problem. What we do solve is the record-keeping infrastructure: every figure traceable, every exclusion documented, no last-minute assembly from disconnected spreadsheets.
CSRD Scope 3 reporting is ambitious in its scope by design. The intent is to make value-chain emissions visible, not just operational ones. That is the right direction. The practical challenge is that year one requires more assessment work than most companies anticipate, and the documentation of boundary decisions matters as much as the numbers themselves.