Scope 3 Emissions Reporting Is Now Mandatory: Here's How to Actually Collect the Data

We've talked to about forty companies since November about their Scope 3 readiness. The honest summary: most have done nothing. Not because they don't care. Because they looked at the fifteen GHG Protocol categories, the supplier data requirements, the estimation methods — and froze.

That's a rational response, by the way. Scope 3 is genuinely hard. It's the part of carbon accounting where tidy formulas meet messy reality. Your Scope 1 is combustion you control. Your Scope 2 is electricity you buy. Your Scope 3 is... everything else. Every supplier, every business flight, every tonne of concrete that arrived on site, every product you sold and what happened to it after.

But scope 3 emissions reporting is now mandatory in Australia for Group 1 entities entering their second ASRS reporting year. And Group 2 entities — mid-market companies captured from 1 July 2026 — get a first-year deferral, but only a fool would wait.

This article is about doing the work. Which categories to prioritise. How to actually get the data. Where estimates are acceptable. And where you're kidding yourself.

The Deferral Clock Is Running

AASB S2 gave every reporting group a one-year exemption from Scope 3 disclosure. That was deliberate — the standard-setters knew this was the hard part.

For Group 1 entities (those with $500M+ revenue, $1B+ gross assets, or 500+ employees), the clock started ticking for financial years beginning on or after 1 January 2025. Their first-year reports — due alongside annual financials in early 2026 — could skip Scope 3. Their second-year reports can't. If your financial year starts 1 July 2025, Scope 3 is mandatory in FY27. If it starts 1 January 2026, it's mandatory in calendar year 2027. Either way, the data collection needs to happen now.

Group 2 entities begin reporting for financial years from 1 July 2026. They get the same one-year Scope 3 deferral — meaning Scope 3 becomes mandatory in their FY28 reports. That sounds far away. It isn't. Getting Scope 3 data from suppliers takes twelve to eighteen months of relationship-building and system setup. Starting that process in mid-2027 is too late.

Here's the part that should matter to your directors: Scope 3 sits inside the modified liability framework. For the first three years (1 January 2025 to 31 December 2027), only ASIC can take enforcement action on Scope 3 disclosures — private litigants can't. After that, full exposure. But "only ASIC" is not "no one." And your Scope 1 and 2 numbers carry full director liability from day one. Getting those right while ignoring Scope 3 is like fixing the roof while the basement floods.

Fifteen Categories. You Don't Need All Fifteen

The GHG Protocol's Corporate Value Chain Standard defines fifteen Scope 3 categories. AASB S2 requires you to consider all of them and disclose which ones you've included. But — and this matters — the standard doesn't require you to report on categories that aren't material to your business.

The "reasonable and supportable information available without undue cost or effort" principle in AASB S2 is your friend here. It explicitly permits estimates. It acknowledges that data quality will improve over time. But it's not a blanket excuse to skip the hard categories just because they're hard. If purchased goods and services make up 60% of your value chain emissions, you can't hand-wave it away as "too difficult."

So which categories matter most? It depends on your industry, but patterns are clear.

Construction and infrastructure: Category 1 (purchased goods and services) dominates. Concrete, steel, timber, rebar — these embodied carbon numbers dwarf everything else. It's common for Category 1 alone to represent over 70% of a construction company's total Scope 3. Category 4 (upstream transport and distribution) and Category 5 (waste generated in operations) are typically next. Business travel and employee commuting exist but are rounding errors by comparison.

Property management and REITs: Category 13 (downstream leased assets) is often the big one — the energy consumption of your tenants. Category 1 and Category 2 (capital goods, including building materials in development projects) matter for companies doing development. Category 5 (waste) is relevant but usually smaller. If you're a pure landlord collecting rent, your Scope 3 story is primarily about tenant emissions.

Professional services and office-based businesses: Category 1 (purchased goods and services, including IT equipment and cloud services), Category 6 (business travel), and Category 7 (employee commuting) tend to be the material ones. Category 1 is still usually the largest, but the gap between categories is narrower than in heavy industry.

Manufacturing: Category 1 again, plus Category 4 (upstream transport), Category 9 (downstream transport and distribution), and Category 11 (use of sold products). If you make something that burns fuel or uses electricity during its lifetime — like machinery, vehicles, or appliances — Category 11 can be enormous. An equipment manufacturer we talked to found that use-of-sold-products emissions were four times larger than all other Scope 3 categories combined.

The practical advice: run a screening assessment first. Use spend-based estimates to get an order-of-magnitude view across all fifteen categories. It'll be rough. That's fine. The point is to find which three or four categories matter and then put your effort there.

Spend-Based vs Activity-Based: An Honest Comparison

This is where people get confused, so we'll be blunt.

Spend-based method: Take your procurement spend in dollars, multiply by an emission factor expressed in kg CO2-e per dollar. The factors come from environmentally-extended input-output (EEIO) databases. It's fast. You can do it with your existing accounts payable data. But the accuracy is terrible — we're talking uncertainty ranges of ±50% or more. A 2022 study by Steubing et al. found that more than half of product footprints calculated from EEIO data differed from life cycle assessment results by a factor of two. For mining-related procurement, the discrepancy was 43%. For waste, 90%.

Here's why that matters practically. Say you spend $2M on concrete. The spend-based factor gives you a number. But that number doesn't distinguish between concrete with 30% fly ash replacement (lower carbon) and standard GP concrete (higher carbon). It doesn't care whether the supplier runs their kiln on gas or coal. It doesn't know if the concrete came from a plant 5km away or 500km away. It treats every dollar of concrete spend identically.

Activity-based method: Collect actual physical data — tonnes of material, litres of fuel, kWh of energy, kilometres travelled — and multiply by emission factors specific to that activity. Far more accurate. Far harder to get. It requires supplier engagement, data sharing agreements, and often physical measurement.

Supplier-specific method: The gold standard. Get the actual carbon footprint data from the supplier for the specific product they sold you. Requires your supplier to have done their own emissions accounting. In Australia in 2026, maybe 10-15% of large suppliers can provide this. For SME suppliers, it's close to zero.

Here's our take, and not everyone agrees with this: start with spend-based for everything, then upgrade the categories that matter.

A spend-based screening across all fifteen categories might take a week. An activity-based calculation for your top three categories might take three months. But without the screening, you don't know which three categories deserve those three months. A Sphera survey from 2025 found that 65% of companies now use two or more data types — up from 48% in 2024. The hybrid approach is becoming the norm, and for good reason.

The Supplier Data Problem Nobody Wants to Talk About

A 2025 MIT survey of 1,200 professionals across 97 countries found that 70% cited lack of available supplier data as their most significant Scope 3 challenge. Not methodology. Not cost. Just getting the data at all.

We're not sure this gets better quickly. And we should be honest about that.

The theory goes like this: you send your suppliers a questionnaire asking for their emissions data. They fill it in. You plug it into your model. Done. The reality looks more like this: you send 200 suppliers a questionnaire. Maybe 40 respond. Of those 40, maybe 15 give you usable numbers. The rest provide data in inconsistent formats, use different reporting boundaries, or just say "we don't track that." One survey response we reviewed had a supplier reporting their "carbon footprint" as a single number with no methodology, no boundary description, and no time period. Was it Scope 1 only? Scope 1 and 2? Did it include their Australian operations or their parent company in Germany? Nobody knew.

And then there's the scale problem. 71% of respondents in the Sphera 2025 Scope 3 Report said they have too many suppliers to engage meaningfully. A mid-sized construction company might have 500+ active suppliers in a given year. Sending all of them a bespoke emissions questionnaire isn't realistic.

What actually works, from what we've seen:

Tier your suppliers by spend. Your top 20 suppliers by dollar value probably account for 60-80% of your Category 1 emissions. Engage those 20 properly. For the remaining 480, use spend-based estimates and improve the data over time.

Use industry averages where supplier data doesn't exist. The NGA Factors workbook from DCCEEW covers some upstream emission factors. For construction materials, Environmental Product Declarations (EPDs) provide product-specific numbers that are far better than EEIO factors even if they're not supplier-specific. The Australian Life Cycle Inventory Database is another source, though it hasn't been updated as frequently as we'd like.

Ask for what you can get. Instead of a 40-field emissions questionnaire, ask your key suppliers for three things: their total annual Scope 1 and 2 emissions, their annual revenue, and the percentage of their revenue that your spend represents. That's enough to calculate an allocated footprint. It's not perfect. But it's a massive improvement over spend-based estimates.

Set a deadline. Suppliers respond to deadlines tied to commercial relationships. "We need this data by March to comply with our mandatory climate reporting" hits differently than "we're working on a sustainability initiative." Especially if procurement sends the email, not the sustainability team.

What "Good Enough" Looks Like in Year One

Here's where we'll say something that might irritate the purists: your first Scope 3 disclosure will be wrong. Not slightly imprecise. Materially incomplete. Based on estimates that might be off by 30-50% in some categories.

That's okay. The standard expects this.

AASB S2 requires you to use "all reasonable and supportable information available at the reporting date without undue cost or effort." That phrase does real work. It means your assurance provider won't fail you for using spend-based estimates in categories where activity data doesn't exist yet. It means you can disclose that Categories 8, 10, 14, and 15 are not material to your business and exclude them. It means your data quality can — and should — improve year on year.

What "good enough" looks like in practice:

A completed screening assessment across all 15 categories, even if most use spend-based estimates
Activity-based or supplier-specific data for your top 2-3 material categories
Clear disclosure of which categories you've included and which you've excluded (and why)
Documentation of your methodology, data sources, and known limitations
A plan — written down, not just discussed — for improving data quality in year two

What's not good enough:

Skipping the screening entirely and reporting only on easy categories like business travel and employee commuting
Using spend-based estimates for categories you know are material and could get better data for
Reporting a single number with no category breakdown
Disclosing Scope 3 only because you have to, with no intention of using the data to actually reduce emissions

That last point isn't just about compliance. The ACCC is actively pursuing greenwashing cases — Clorox copped an $8.25 million penalty in February 2025 for misleading environmental claims. If you report Scope 3 numbers in your annual report but make net-zero claims elsewhere that your own data contradicts, you're in ACCC territory. Your Scope 3 disclosure needs to be consistent with every other environmental claim your company makes.

The Tool Question

We build carbon accounting software, so take this with the appropriate grain of salt. But the tool landscape for Scope 3 is genuinely underdeveloped compared to Scope 1 and 2.

Scope 1 and 2 calculations are formulaic. You know the inputs, you know the factors, you know the output. We can automate extraction of energy consumption from utility bills because the underlying maths is clean.

Scope 3 is different. The data comes from dozens of sources in dozens of formats. Procurement systems. Travel booking platforms. Waste management reports. Freight invoices. Supplier questionnaire responses in Excel, PDF, email text, and once — I'm not making this up — a handwritten fax. No single tool handles all of this elegantly yet.

What software can do well: maintain the emission factor libraries (EEIO databases, NGA Factors, industry-specific factors), run the spend-based calculations at scale once you feed in procurement data, track supplier responses and flag gaps, and produce the AASB S2-compliant disclosures with category breakdowns and methodology documentation.

What software can't do for you: make your suppliers respond to data requests, validate whether a supplier's self-reported emissions number is accurate, or decide which categories are material to your business. Those are human judgment calls. They require people who understand your operations, your supply chain, and your industry.

We're building Scope 3 capabilities into Carbonly.ai — spend-based screening, supplier data collection workflows, category-level reporting. But we won't pretend it's a push-button solution. Nobody's is. Anyone who tells you they've "solved" Scope 3 is selling you something that'll fall apart the first time an auditor tests it.

Start Now. Seriously

If you're a Group 1 entity, your Scope 3 deferral is gone or almost gone. You should be collecting data right now for the reporting period you're in. Not planning to collect data. Actually collecting it. Sending supplier questionnaires. Pulling procurement spend reports. Running screening assessments.

If you're a Group 2 entity captured from July 2026, you've got one year of Scope 3 grace. Use that year to run the screening, tier your suppliers, and send the first round of data requests. When mandatory Scope 3 hits in your second year, you'll have twelve months of supplier engagement behind you instead of zero.

One specific action, right now: export your top 50 suppliers by annual spend from your procurement or AP system. Map each to a GHG Protocol Scope 3 category. Run a spend-based estimate using EEIO factors (the US EPA supply chain factors are publicly available and widely used as a starting point; Australian-specific EEIO data is thinner, which is a problem the industry hasn't solved). That exercise alone will tell you where 80% of your Scope 3 sits. It'll take half a day. And it'll replace the paralysis with a plan.

Related reading: