Scope 3 Emissions: Complete Guide for Enterprises 2026

Most Scope 3 conversations start with the wrong question. Teams ask “how do we calculate this?” before they’ve settled a more basic one: which of the fifteen categories actually matter for this business, and which regulatory clock is really running against them this year? Scope 3 is not one number — it is a portfolio of value-chain emissions sources that behave differently by industry, and the 2026 regulatory landscape treats them differently depending on company size and jurisdiction.

Why does Scope 3 dwarf Scope 1 and 2 for most companies?

Scope 1 covers direct emissions from sources a company owns or controls — fuel burned in a company vehicle, gas burned in a boiler. Scope 2 covers indirect emissions from purchased electricity, heat, and steam. Scope 3 covers everything else across the value chain: everything a company buys, everything it ships, everything employees do to get to work, and — for many companies — everything that happens to a product after it’s sold. For most organizations outside heavy industry, Scope 3 is the largest share of total emissions by a wide margin, often several times larger than Scope 1 and 2 combined. That is precisely why regulators have started requiring it even as they continue to debate exactly how strictly.

What are the 15 GHG Protocol categories?

Which three categories usually matter most — and why?

Building detailed data pipelines for all fifteen categories before knowing which ones are material is the single most common wasted effort in a first Scope 3 program. In practice, three categories account for the overwhelming majority of Scope 3 emissions across most industries.

Category 1: purchased goods and services

For manufacturers and retailers, Category 1 is typically the largest single Scope 3 category, often accounting for half or more of total value-chain emissions. A furniture manufacturer’s largest exposure is usually timber, foam, and textile inputs; a retailer’s largest exposure is usually the manufacturing footprint embedded in the products it resells rather than anything it does directly. This is also the category where the primary-versus-secondary data distinction matters most in practice, since spend-based estimates across thousands of SKUs can be wildly imprecise compared to supplier-specific figures for even a handful of high-volume product lines.

Category 11: use of sold products

For consumer electronics, automotive, and appliance manufacturers, a product’s operating life frequently outweighs its manufacturing footprint. A car sold today will emit far more over ten years of driving than it did rolling off the production line; a data center server will draw more lifetime electricity than the emissions embedded in manufacturing it. Companies in these sectors that focus exclusively on their factory floor while ignoring Category 11 are typically missing the majority of their real footprint.

Category 15: investments

For banks, insurers, and asset managers, Category 15 — financed emissions across a loan and investment portfolio — routinely dwarfs the institution’s own operational footprint by one or two orders of magnitude. This category follows PCAF (Partnership for Carbon Accounting Financials) methodology, which assigns a data-quality score from 1 (audited, company-reported data) to 5 (fully modeled estimates) and requires different attribution approaches for corporate loans, listed equity, and project finance. An institution reporting a single blended Category 15 figure without breaking out asset class and data quality tier is producing a number that will not survive serious scrutiny from an assurance provider or a sophisticated ESG analyst.

What about the other twelve categories — do they ever matter?

The remaining categories are rarely a company’s largest single source, but ignoring them entirely is its own risk, since an auditor or assurance provider will expect a documented reason for excluding any category, not silence.

Category 2 (capital goods) matters most in years of heavy investment — a manufacturer building a new plant or a data center operator installing new server halls will see a temporary spike here that should not be mistaken for a trend in operational emissions.

Category 3 (fuel- and energy-related activities) captures the upstream emissions of producing the fuel and electricity a company already counts in Scope 1 and 2 — commonly a modest single-digit percentage addition, but one regulators increasingly expect to see rather than omit.

Category 4 (upstream transportation and distribution) can be significant for companies sourcing heavy or bulky inputs across long distances — a mining or heavy-materials business will see this category weigh more than a services company ever will.

Category 5 (waste generated in operations) is usually small in proportion but easy to measure well, since most companies already track waste volumes for cost or regulatory reasons unrelated to carbon.

Category 6 (business travel) and Category 7 (employee commuting) are rarely material at enterprise scale but are highly visible to employees and are often the first categories a company measures, since the data — expense reports, commuting surveys — is the easiest to access.

Category 8 (upstream leased assets) applies narrowly: assets a company leases in but does not already count under Scope 1 or 2, most relevant to companies leasing significant equipment or vehicle fleets.

Category 9 (downstream transportation and distribution) matters most for companies that sell through independent distributors or retailers who arrange their own outbound logistics — the reporting company pays for none of this freight directly, but the Protocol still requires it in the inventory.

Category 10 (processing of sold products) is a business-to-business category: relevant mainly to companies selling intermediate materials — resins, metals, chemical intermediates — that a customer will further process before a final product exists.

Category 12 (end-of-life treatment) matters most for packaging-heavy consumer goods and durable products with a defined disposal pathway, and is an increasing focus of extended producer responsibility regulation independent of climate disclosure.

Category 13 (downstream leased assets) is the mirror of Category 8: relevant to companies that lease out property or equipment they own, such as a commercial landlord or an equipment-rental business.

Category 14 (franchises) is material almost exclusively for franchisors — a restaurant or retail franchise brand must account for emissions from franchisee-operated locations it does not directly control but does financially benefit from.

What does a Scope 3 program look like by industry?

What does supplier engagement actually involve?

Most Scope 3 programs discover the same pattern: a small share of suppliers accounts for the large majority of spend and, by extension, the large majority of Category 1 emissions. Segmenting suppliers by spend before launching any data request is the difference between a program that produces useful data and one that produces a low response rate and little usable improvement.

A typical engagement escalation runs in stages: start with a spend-based estimate for the full supplier base, layer in activity-based data (volumes, quantities) for mid-tier suppliers where it is readily available, and reserve direct primary-data requests — surveys, verified environmental product declarations, or product carbon footprints — for the top tier of suppliers by spend or emissions contribution. Response rates to direct supplier data requests vary enormously by sector and by the strength of the commercial relationship; a supplier that depends heavily on the requesting company will typically respond faster than one selling into a fragmented customer base with little individual leverage.

How do emission factor databases work, and which one should you trust?

An emission factor converts an activity or spend amount into an estimated quantity of CO2e. Publicly maintained databases — such as those published by national environmental agencies, the GHG Protocol’s own supporting tools, and commercial life-cycle-assessment databases — provide factors by region, sector, and material type. No single database is universally “correct”; the practical requirement under CSRD and SBTi alike is not that a company picked the single best factor, but that it can document which database it used, why, and how consistently it applied that choice across reporting periods.

What does CSRD actually require in 2026?

Following the Omnibus simplification (Directive EU 2026/470, effective 19 March 2026), CSRD now applies to companies with more than 1,000 employees and more than €450 million in net turnover; listed SMEs are exempt. This represents a substantial narrowing from CSRD’s original scope. Where climate is assessed as material under double materiality — which auditors treat as the default assumption for most sectors — in-scope companies disclose gross Scope 3 emissions by category under ESRS E1, alongside the methodology and the primary-versus-secondary data split used for each category.

What does California’s SB 253 require?

SB 253 applies to companies with more than $1 billion in annual revenue doing business in California, regardless of headquarters location. Scope 1 and 2 disclosures are due first, with Scope 3 reporting following in the subsequent filing cycle, covering the prior fiscal year’s data. Assurance requirements phase in gradually — starting with limited assurance on Scope 1 and 2, with the specific assurance requirements for Scope 3 still under the California Air Resources Board’s (CARB) rulemaking process as of 2026.

What does SBTi require — and is it mandatory?

The Science Based Targets initiative (SBTi) is voluntary, but it has become the de facto benchmark investors, lenders, and large customers use to judge whether a company’s climate targets are credible. SBTi requires a Scope 3 near-term target for any company where Scope 3 represents 40% or more of total emissions — a threshold most companies exceed once a complete inventory is built.

Following the Corporate Net-Zero Standard V2 (published June 2026), near-term Scope 3 targets now focus on “significant categories” — generally those representing more than 5% of total Scope 3 emissions — rather than requiring equal depth of ambition across all fifteen categories. This shift matters in practice: it formally recognizes what most experienced sustainability teams already do informally, which is concentrate rigor on the two or three categories that actually drive the number.

Primary data vs secondary data: which do regulators actually expect?

Not all Scope 3 data is equal, and regulators increasingly ask companies to say which is which. Secondary data — industry-average emission factors applied to spend or activity volumes — is the fastest way to produce a first complete inventory but carries the least precision. Primary data — emissions reported directly by a supplier, or figures from a verified environmental product declaration — is more defensible under assurance but far slower to collect at scale.

The practical sequencing most sustainability teams follow: start with spend-based secondary data to establish a complete baseline across all fifteen categories, confirm which two or three are material, and only then invest in primary data collection for those specific categories. Chasing primary data across all fifteen categories from day one is rarely a good use of a first year’s budget.

How mature does a Scope 3 program need to be, stage by stage?

Stage 1 — no program. No Scope 3 estimate exists yet. The first step is a spend-based screening across all fifteen categories, using industry-average emission factors, purely to identify which categories are material for this specific business.

Stage 2 — basic program. A spreadsheet-based estimate exists but lives with one person and is rebuilt from scratch each year. The priority at this stage is documenting the methodology and emission-factor sources used, so the estimate can be defended and repeated consistently.

Stage 3 — governed program. Ownership of Scope 3 data collection is assigned, and the organization has identified its two or three material categories. The priority shifts to activity-based or supplier-specific data for those categories, while documenting exclusions for the rest with quantitative justification rather than silence.

Stage 4 — mature program. The inventory supports external assurance. A base year is locked, the primary-versus-secondary data split is documented by category, and every calculation carries a traceable formula and emission-factor source suitable for an auditor to test.

Stage 5 — leading program. Scope 3 data feeds directly into supplier engagement, procurement decisions, and product design, not just annual disclosure. The organization actively works with suppliers to shift from secondary to primary data over time, rather than treating the inventory as a compliance exercise repeated annually with no underlying change.

What does an assurance-ready Scope 3 inventory need to demonstrate?

What glossary terms should a Scope 3 report use correctly?

What mistakes derail a Scope 3 inventory under assurance?

Treating a data gap as grounds for silent exclusion. A missing category requires a documented plan to address it, not an unexplained omission — auditors and assurance providers treat undocumented gaps far more harshly than a disclosed, justified estimate.

Misclassifying spend-based estimates as primary data. Activity-based data using industry-average emission factors is secondary data under ESRS E1’s own definitions. Labeling it as primary inflates a company’s reported data quality and creates unnecessary assurance risk once auditors test the underlying source.

Changing methodology after locking a base year. Both SBTi and CSRD expect a stable base year and consistent methodology. Switching calculation approaches mid-stream creates a second, avoidable restatement obligation across regulatory and voluntary disclosures alike.

Reporting a single blended figure for a category that spans multiple sub-populations. This is most visible in Category 15 for financial institutions, where a single financed-emissions number without asset-class and data-quality breakdowns will not survive scrutiny from a sophisticated analyst or auditor — but the same logic applies to Category 1 for a manufacturer with wildly different supplier types.

What’s a realistic timeline for a first defensible Scope 3 inventory?

Most organizations starting from Stage 1 can expect roughly the following arc, though complex multi-entity organizations should expect the upper end of each range. In the first one to two months, a spend-based screening across all fifteen categories identifies which two or three are material. Over the following two to four months, the organization documents its methodology, assigns internal ownership, and begins activity-based data collection for the identified priority categories. By month six to nine, primary data requests to top-tier suppliers for the priority categories are underway, alongside documentation of exclusions for immaterial categories. By month nine to twelve, the inventory is complete enough to support a first external disclosure, though typically not yet at the assurance-ready standard that Stage 4 requires — that level of rigor generally takes a full second reporting cycle to reach, since it depends on having at least one prior year’s data to compare against and restate consistently.

Frequently asked questions

Is Scope 3 mandatory for every company under CSRD?

Only for companies within CSRD’s scope following the Omnibus revision — generally those with more than 1,000 employees and more than €450 million in turnover — and even then, only where climate is assessed as material under double materiality, which auditors treat as the default for most sectors.

Which Scope 3 category should a company measure first?

Whichever category is most likely to dominate for that industry: purchased goods and services for manufacturers and retailers, use of sold products for consumer electronics and automotive, and investments for financial institutions. A spend-based screening across all fifteen categories is the fastest way to confirm which ones actually apply.

Does SBTi validation satisfy CSRD’s Scope 3 disclosure requirement?

Not automatically. The two frameworks are complementary rather than interchangeable — SBTi validates the ambition of a target, while CSRD requires a disclosed, auditable inventory and transition plan. A well-structured Scope 3 inventory can support both, since both frameworks rely on the same underlying GHG Protocol data.

What happens if a company can’t get supplier-level data for a material category?

Spend-based or activity-based secondary data is an accepted starting point under both CSRD and SBTi, provided the company documents its methodology and shows a credible plan to improve data quality over time — silence about a data gap is treated far more harshly than a documented estimate.

How does Scope 3 differ for a bank versus a manufacturer?

For a manufacturer, Scope 3 is dominated by purchased goods and, depending on the product, use-phase emissions. For a bank or asset manager, Scope 3 is dominated by Category 15 (investments) — financed emissions calculated under PCAF methodology across the loan and investment portfolio, which typically dwarfs the institution’s own operational footprint.

How many of the fifteen categories does a typical company actually need to report in detail?

Most organizations find that two or three categories account for the large majority of their Scope 3 footprint. SBTi’s Net-Zero Standard V2 formally reflects this by focusing near-term targets on categories above a 5% materiality threshold rather than demanding equal rigor across all fifteen.

What’s the difference between the value chain cap and full Scope 3 exemption for smaller suppliers?

The value chain cap under CSRD/ESRS E1 limits what a large in-scope customer can demand from a smaller supplier — generally no more than the Voluntary SME Standard already requires — but it does not exempt that supplier from its own separate regulatory obligations if it happens to cross a different threshold elsewhere.

Should a company wait for final SB 253 Scope 3 rules before starting its inventory?

No. Building a GHG Protocol-compliant inventory now serves CSRD, SBTi, and SB 253 simultaneously, since all three rely on the same underlying category structure. Waiting for final assurance rules to be settled before starting data collection typically costs more time than it saves.

Is it acceptable to use different emission factor databases for different categories?

Yes, provided the choice is documented and applied consistently across reporting periods for each category. What undermines credibility is switching sources between reporting years without disclosure, which can make emissions appear to change for reasons unrelated to actual performance.

Does a small supplier need its own full Scope 3 inventory to respond to a customer’s data request?

Not necessarily. Under the CSRD value chain cap, a supplier with fewer than 1,000 employees cannot be compelled to provide more than the Voluntary SME Standard requires, even if the requesting customer is itself fully in scope for CSRD.

What’s the single most common reason a Scope 3 inventory fails assurance?

Undocumented category exclusions and undisclosed methodology changes are the two most frequent causes — both are avoidable with disciplined documentation from the very first reporting cycle, rather than something that can be fixed retroactively once an auditor asks the question.

Where to go next

Building a first Scope 3 inventory usually surfaces a data-coordination problem before it surfaces an audit-readiness one. Organizations comparing platforms for this stage often start with Sweep for value-chain data coordination or Persefoni for audit-grade Scope 1–3 ledgers, depending on which problem is more urgent — see Sweep vs Persefoni compared for the fuller breakdown. For the regulatory background behind CSRD’s current scope, see What is CSRD?