The 15 Scope 3 Categories: Which Ones Actually Matter for Your Business

CDP estimates that Scope 3 emissions account for 75% of the average company's total greenhouse gas footprint. For financial services firms, it's 99.98%. For Australian retailers sourcing from Asia, it's north of 90%. And yet, when we ask companies which of the 15 scope 3 categories apply to them, the most common answer is a blank stare.

This isn't ignorance. It's overwhelm. The GHG Protocol's Corporate Value Chain Standard lists fifteen categories of indirect emissions, split between upstream and downstream. AASB S2 paragraph 29(a)(vi) requires you to disclose which categories you've included. That means you can't just report a single Scope 3 number — you need to show your working, category by category.

But here's what the standard also says, and what most guidance documents bury in footnotes: you only need to report on categories that are material to your business. For most Australian companies, that means 5 to 8 categories. Not fifteen. The trick is knowing which ones.

We built Carbonly to track all 15 Scope 3 subcategories per the GHG Protocol. Every emission entry tagged as Scope 3 gets assigned to a specific category from a dropdown — Category 1 through 15. That data structure exists so your reports can break emissions down by subcategory, which is exactly what auditors and AASB S2 demand. But the system also shows, very clearly, that most companies concentrate their data in just a handful of those categories.

Here's every category. What it covers, who it matters for, and how hard the data is to get.

Upstream Categories (1-8): Your Supply Chain

The first eight categories cover emissions that happen before your product or service reaches your customer. This is where most of the work lives.

Category 1 — Purchased Goods and Services is almost always the biggest. It covers the cradle-to-gate emissions of everything you buy: raw materials, components, office supplies, professional services, cloud hosting. CDP data shows that Category 1 and Category 11 (use of sold products) together represent 84% of all reported Scope 3 emissions globally.

For a construction company, Category 1 is concrete, steel, and aluminium — materials with emission factors of roughly 0.2, 2.2, and 20 kg CO2-e per kilogram respectively. For a retailer, it's the embodied emissions of every product on the shelf. For a professional services firm, it's surprisingly small because you're not buying physical goods.

The data challenge: this is the hardest category to measure accurately. Supplier-specific emission data barely exists in Australia — maybe 10-15% of your top suppliers can provide it today. The fallback is spend-based estimation, where you multiply procurement spend by an industry-average emission factor from databases like EXIOBASE or Climatiq. Research comparing EXIOBASE with process-based lifecycle databases like ecoinvent has found that the two methods can differ by a factor of two for the same product. Spend-based gets you a starting number. Don't confuse it with an accurate one.

Carbonly supports both approaches. For spend-based, you can enter amounts in AUD, USD, EUR, GBP, or NZD — the system handles currency conversion automatically and applies EXIOBASE or Climatiq factors. For activity-based, you enter physical quantities and match them against specific emission factors from our material library, which draws on NGA, BEIS, IEA, IPCC, Ecoinvent, EPIC, and FootprintLab. Our AI document extraction can also pick up monetary amounts from supplier invoices directly, which speeds up the spend-based approach significantly.

Category 2 — Capital Goods covers the emissions embodied in assets you purchase: buildings, vehicles, heavy machinery, IT equipment, furniture. The logic is the same as Category 1 — you're counting the cradle-to-gate emissions of the asset — but these are long-lived items rather than consumables.

This matters most for capital-intensive industries. A construction company buying a fleet of excavators, a property developer purchasing building materials for a new development, a mining company commissioning a processing plant. For a consulting firm buying some laptops and chairs, it's probably immaterial.

The data challenge: emission factors for capital goods are patchy. You can get reasonable data for vehicles and standard equipment, but custom-built industrial equipment? Good luck. Most companies start by estimating capital goods emissions using spend-based factors and then refining for the big-ticket items where specific lifecycle data exists (like vehicles, which have well-documented embodied carbon figures of 8-12 tonnes CO2-e per unit for passenger cars).

Category 3 — Fuel- and Energy-Related Activities is the category that confuses everyone. It covers the upstream emissions from the production and transportation of fuels and electricity you consume — but only the parts not already counted in Scope 1 or Scope 2.

Think of it this way. When you burn natural gas (Scope 1), you count the emissions from combustion. Category 3 counts the emissions from extracting that gas, processing it, and piping it to your site. When you buy electricity (Scope 2), you count the emissions from generating it. Category 3 counts the transmission and distribution losses, and the upstream emissions of the fuel used in the power station.

The NGA Factors 2025 workbook actually provides Scope 3 emission factors for purchased electricity alongside the Scope 2 factors. NSW is 0.03 kg CO2-e/kWh for Scope 3, Victoria is 0.09, and the national average is 0.07. These are small relative to the Scope 2 factors (0.64 and 0.78 respectively) but they're not zero, and for electricity-heavy businesses they add up.

This category is relevant for every company that reports Scope 1 or 2. It's also where double-counting risk creeps in — if a fuel supplier's transport emissions are in your Category 3, make sure you're not also counting them in Category 4 (upstream transport). The GHG Protocol designed the categories to be mutually exclusive, but the boundaries aren't always obvious in practice.

Category 4 — Upstream Transportation and Distribution covers emissions from transporting purchased goods from suppliers to your operations. The key distinction: this only includes transport you don't directly pay for. If you hire a freight company, those emissions go in Category 4. If you own the trucks, it's Scope 1.

Australian geography makes this category material for almost everyone. A Brisbane manufacturer sourcing steel from Wollongong, a Perth retailer receiving goods from Melbourne, a Darwin construction company importing equipment from Southeast Asia. Distances are massive and freight is carbon-intensive.

Data can come in tonne-kilometres, vehicle-kilometres, or passenger-kilometres — Carbonly supports all three transport-specific units. The challenge is that many Australian businesses don't know the transport mode or distance for every inbound shipment. Your supplier delivers to your site, you get an invoice for goods, and the logistics are invisible to you. In practice, you'll need to work with your procurement team to map major supply routes.

Category 5 — Waste Generated in Operations counts the emissions from treating and disposing of waste your business produces. Landfill methane, incineration emissions, recycling process emissions. Not the waste from your products after sale (that's Category 12) — just what comes out of your own operations.

For office-based companies, this is usually immaterial. For manufacturing, food processing, and construction, it can be significant. A single large landfill generates methane with a warming impact 28 times that of CO2. But for most businesses, waste is 1-3% of their total footprint. The data is usually available from your waste contractor — tonnage by waste stream — and the emission factors come from DCCEEW's NGA Factors.

Category 6 — Business Travel is one of the easiest categories to measure and one of the most politically sensitive. Flights, hotels, hire cars, trains. PwC found that 48% of professional services emissions come from business travel. And cabin class matters enormously — a business-class seat generates roughly 3-4 times the emissions of economy because of the floor space it occupies.

Travel management company (TMC) data is your best source here. If your company uses a corporate travel provider, they can usually export booking data with distances and cabin classes. Carbonly supports accommodation-specific units like room-nights alongside the standard transport units. For companies without a TMC, corporate credit card data classified by merchant category is the next best option.

Category 7 — Employee Commuting covers emissions from staff getting to and from work. It's typically estimated using surveys — asking employees how far they travel and by what mode — combined with emission factors per passenger-kilometre. Carbonly uses FTE and square-metre-based estimation methods where survey data isn't available.

Honestly? For most companies, this is a minor category. A 200-person office might generate 200-400 tonnes CO2-e from commuting annually. It's worth measuring once, noting the methodology, and updating every couple of years unless your workforce size or location changes dramatically.

Category 8 — Upstream Leased Assets covers emissions from assets you lease but don't include in your Scope 1 or 2. This is relevant mainly if you use the financial control approach for your emissions boundary rather than operational control. Under NGER's operational control test — which is what most Australian companies use — leased assets where you have operational control are already in Scope 1 and 2. So for most Australian reporters, Category 8 is not applicable.

If you do use financial control as your consolidation approach, this is where leased office space, leased vehicles, and leased equipment emissions sit. But double-check your boundary approach before spending time on this one. We've seen companies measure Category 8 when it was already captured in their direct scopes. That's double counting, and it inflates your numbers.

Downstream Categories (9-15): After the Sale

These categories cover what happens to your products and services after they leave your hands. They're most material for manufacturers, energy producers, and financial institutions.

Category 9 — Downstream Transportation and Distribution is the mirror of Category 4. It covers transport emissions between your company and the customer — but only for transport you don't pay for or control. If a distributor picks up goods from your warehouse and delivers them to retailers, and you don't pay the freight bill, those emissions are Category 9.

This matters for manufacturers and wholesalers with third-party distribution networks. For a services company, it's almost always immaterial.

Category 10 — Processing of Sold Products applies when you sell intermediate products that undergo further processing before reaching the end user. A steel manufacturer selling coils to an automotive company. A flour miller selling to a bakery. The emissions from the downstream processing — not the use of the final product, just the processing step — go here.

Most Australian companies outside heavy manufacturing can screen this out.

Category 11 — Use of Sold Products is the other giant category. For energy companies, it's everything. When an Australian gas producer sells LNG, the emissions from burning that gas are Category 11. For BHP, the emissions from steelmaking with their iron ore. For a car manufacturer, the lifetime fuel consumption of every vehicle sold. CDP data shows Category 11 is the single largest Scope 3 category for fossil fuel producers, often exceeding all other categories combined.

For non-energy businesses, the materiality depends on whether your product consumes energy during use. Selling electrical appliances? Category 11 matters. Selling clothing? Probably not, unless you count the washing and drying (some companies do, most don't for apparel).

Category 12 — End-of-Life Treatment of Sold Products covers the emissions from disposing of your products when the customer is done with them. Landfill methane from food packaging. Incineration emissions from electronics. For companies selling products with significant end-of-life emissions — packaging manufacturers, food companies, electronics brands — this can be material.

For services companies, it's not applicable.

Category 13 — Downstream Leased Assets is the landlord category. If you own buildings or equipment and lease them to tenants, the tenants' Scope 1 and 2 emissions from operating those assets are your Category 13. This is big for property managers and REITs. A commercial landlord with 50 tenanted buildings has significant Category 13 emissions from tenant electricity use, gas consumption, and refrigerant leakage.

The data challenge here is that tenants often don't share energy data with their landlord. Some lease agreements include green clauses requiring energy disclosure, but enforcement is patchy.

Category 14 — Franchises covers emissions from franchise operations. If you're a franchisor — think quick-service restaurant chains, fuel retailers, hotel brands — the Scope 1 and 2 emissions of your franchisees go here. If you're the franchisee, this doesn't apply.

Category 15 — Investments is the category that makes financial services companies lose sleep. It covers the emissions attributable to your equity investments, debt investments, and project finance. A bank's financed emissions — the Scope 1 and 2 emissions of every company they lend to, proportioned by the bank's financial exposure — dwarf their own operations by factors of 700 or more.

AASB S2 has a specific disclosure requirement for Category 15: if you include it in your Scope 3 total, you must disclose the Category 15 subtotal separately and break out financed emissions within that. The December 2025 AASB amendments (AASB S2025-1) also provide relief on how financed emissions are classified by industry — a recognition that PCAF data quality scores across entire lending portfolios are still maturing.

For non-financial companies, Category 15 applies only if you hold significant equity investments in other entities. Joint ventures are the common example in Australian mining and resources.

Which Categories by Industry

Rather than listing all fifteen for every sector, here's where we consistently see the concentration for Australian companies.

Construction and infrastructure: Categories 1 (materials, 40-60% of Scope 3), 2 (heavy equipment), 4 (inbound freight), and 5 (site waste). Category 1 alone — concrete, steel, aluminium, timber — dominates. Everything else is secondary.

Property and real estate: Category 13 (downstream leased assets) is the main event for landlords. Category 1 (maintenance materials and services) and Category 6 (travel for property managers with distributed portfolios) are secondary.

Mining and resources: Categories 1 (purchased goods), 4 (transport of inputs), 11 (use of sold products — this is where it gets massive for coal and gas producers), and 15 (for companies with JV interests). Category 11 is often 80-90% of total Scope 3 for fossil fuel producers.

Financial services: Category 15 (investments) accounts for virtually the entire Scope 3 footprint. Everything else is rounding error.

Retail: Categories 1 (purchased goods for resale), 4 (inbound freight), 9 (outbound delivery to customers), and 12 (end-of-life of sold products, especially packaging).

Professional services: Categories 6 (business travel) and 7 (employee commuting) together typically make up 60-80% of Scope 3. Category 1 (purchased services like IT and cloud) makes up most of the rest.

CDP's technical note on Scope 3 relevance by sector confirms this pattern: the two most material Scope 3 categories in each sector cover, on average, 81% of total Scope 3 emissions. You don't need perfection across all fifteen. You need accuracy in the categories that matter.

The Practical Problem: Data Quality

Here's where we stop being theoretical and start being honest about what actually happens when you try to measure these categories.

For Categories 3, 5, 6, and 7, data collection is manageable. Your energy bills give you Category 3 inputs. Your waste contractor gives you Category 5 tonnage. Your travel provider or credit card statements give you Category 6. A staff survey gives you Category 7. These aren't easy, exactly, but the data exists and the emission factors are established.

Category 1 is where it breaks down. Spend-based emission factors from EXIOBASE or Climatiq give you a number, but the uncertainty is significant — research has shown that EXIOBASE-based calculations can differ from process-specific lifecycle assessments by a factor of two or more. And the data input itself — classifying every line item in your accounts payable into the right emission factor category — is grunt work. Carbonly's AI material matching helps here: the system learns which procurement line items map to which emission factors based on whether you're using spend-based or activity-based approaches, and it adapts the matching context accordingly. But there's no way around the initial effort of getting your procurement data structured.

For Categories 9-12, you're estimating based on product characteristics and assumptions about downstream behaviour. How far does your product travel after you sell it? How much energy does it consume during its useful life? What happens to it when it's thrown away? Unless you make bespoke products for a small number of known customers, these are modelled estimates, not measured data.

We're honest about this: Scope 3 data quality in year one will be poor for most companies. The safe harbour under ASRS's modified liability framework protects Scope 3 disclosures from private litigation until 31 December 2027. That buys time to improve. But it won't last forever, and ASIC can still act on Scope 3 statements during the protected period. "Protected" doesn't mean "ignored."

The Screening Process

AASB S2 requires you to consider all fifteen categories. It does not require you to report on all fifteen. The GHG Protocol's Scope 3 Evaluator tool provides a quick screening method, but we'd suggest a simpler approach.

For each category, ask three questions. Does your business have material activity in this area? (If you have no franchises, screen out Category 14.) Can you get data or make a reasonable estimate? And would excluding this category materially misrepresent your total Scope 3 footprint?

Document your screening. Write a paragraph on why each excluded category is immaterial. Auditors will ask. AASB S2 paragraph 29(a)(vi) specifically requires you to disclose which categories you've included — which implicitly means explaining which you haven't.

We've seen companies screen out 6-8 categories legitimately and focus their data collection effort on the 4-6 that actually matter. That focused approach produces better data than spreading thin across all fifteen.

One trap to watch: don't screen out Category 1 just because it's hard. If you buy physical goods, Category 1 is almost certainly material. Spend-based is acceptable in year one. Activity-based is where you're heading by year three. But skipping it entirely because you can't get supplier-specific data? That won't fly with an auditor.

Where Double Counting Hides

The GHG Protocol designed the fifteen categories to be mutually exclusive. In theory, there's no overlap. In practice, we see double counting in three places.

First, Category 3 versus Category 4. The upstream emissions of fuel you consume (Category 3) include emissions from transporting that fuel to you. But if you're also counting inbound freight emissions in Category 4, and some of that freight is fuel delivery, you've counted it twice. The fix: exclude fuel delivery from Category 4 if you're already capturing it in Category 3.

Second, Category 1 versus Category 2. The distinction between purchased goods and capital goods is about asset life, not about the type of item. A laptop used for a year and disposed of could arguably sit in either. Be consistent about your threshold (most companies use the capitalisation threshold from their financial accounting) and document it.

Third, Category 1 versus Category 4. If your suppliers quote you a delivered price (freight inclusive), the transport emissions might be embedded in your spend-based Category 1 calculation. If you then separately estimate Category 4 based on shipping distances, those transport emissions are counted twice. The GHG Protocol acknowledges this risk and recommends either using separate supplier and transport data, or using spend-based for both and accepting the embedded overlap as immaterial.

What Good Looks Like in Year One

Good enough for your first Scope 3 report means: a documented screening across all 15 categories, spend-based estimates for the material categories you can't get activity data for, activity-based calculations where data exists (business travel, waste, fuel-and-energy-related activities), and a clear improvement plan for moving from spend-based to activity-based over the next two to three years.

It doesn't mean: perfect supplier-specific data across your entire value chain. Nobody has that. Not BHP, not Woolworths, not any company on the ASX. Woolworths reports that Scope 3 represents 96% of its total emissions, with supply chain (Category 1) dominating — and even they rely on estimates for most of their 10,000+ suppliers.

Carbonly stores every Scope 3 emission entry against its assigned category, the emission factor source, and the calculation method (spend-based or activity-based). That audit trail is what makes year-on-year comparison meaningful and what gives auditors confidence that your methodology is consistent, even if the data inputs are still improving.

Start with your procurement data for Category 1, your energy bills for Category 3, your waste data for Category 5, and your TMC or expense data for Category 6. That covers the four categories where data is available right now for almost every Australian business. Then map your industry-specific material categories — the table above tells you which ones — and build a twelve-month plan to fill the gaps.

The companies that will be in trouble under ASRS Group 2 reporting from July 2026 aren't the ones with imperfect Scope 3 numbers. They're the ones who haven't started the screening at all.

Related Reading: