Product Carbon Footprinting in Australia: From Corporate Totals to Individual Products

Last month, a mid-sized Australian aluminium fabricator told us their EU customer had given them 90 days to provide product-level carbon footprint data — or lose the contract. Not corporate-level. Not "we've offset our operations." The customer wanted kilograms of CO2-e per tonne of finished product, broken down by lifecycle stage.

That fabricator had a perfectly good corporate emissions inventory. Scope 1, 2, and partial Scope 3, all calculated, all reported. But when a procurement manager in Hamburg asked "what's the carbon footprint of this product?" — none of that helped. A corporate inventory tells you your organisation's total. Product carbon footprinting tells you which of your products carries the most emissions, and more importantly, where in the lifecycle those emissions sit: raw materials, manufacturing, transport, use phase, end of life.

This distinction matters more than it used to. And the gap between "we report our corporate footprint" and "we can tell you the footprint of product X" is where a lot of Australian manufacturers and exporters are stuck right now.

The Question Has Changed

For most of the past decade, carbon reporting meant one thing: corporate-level emissions inventories. Your organisation's total Scope 1, 2, and 3 in tonnes of CO2-e per year. That was enough for NGER reporting, enough for voluntary CDP disclosures, enough for the sustainability section of the annual report.

Then three things happened at roughly the same time.

The EU's Carbon Border Adjustment Mechanism entered its definitive compliance phase on 1 January 2026. Australian exporters of steel, aluminium, cement, fertilisers, and hydrogen now need to provide verified embedded emissions data for those products. EU importers must purchase CBAM certificates at roughly €70-100 per tonne of CO2 — and they won't accept "we don't have product-level data" as an answer. During the initial phase-in, the cost is a fraction of the full price (about 2.5% in 2026), but it ramps to 100% by 2034. And the EU is already proposing to expand CBAM to downstream products.

EPD Australasia hit 3,400 published Environmental Product Declarations from 175 organisations. In the second half of 2025 alone, 27 new companies joined and 1,153 new EPDs were published — mostly for building and construction products. The Green Star Buildings v1.1 rating tool, effective from 1 May 2026, increased the Responsible Product Value for product-specific EPDs from 5 to 7 RPV. That's a direct financial incentive for material suppliers to quantify and publish their product footprints.

And large buyers started asking suppliers for product carbon footprints as a procurement condition. PepsiCo's interim requirement, published May 2025, expects tier-1 suppliers to provide product-level carbon footprint data for major product categories between 2026 and 2028. Woolworths has committed to a 40% FLAG reduction by FY2033 under SBTi — which means they need product-level emissions data from their supply chain, not just corporate totals.

The question has shifted from "what's your organisation's footprint?" to "what's the footprint of the thing you're selling me?"

What Product Carbon Footprinting Actually Involves

A product carbon footprint (PCF) quantifies the greenhouse gas emissions associated with a single product across some or all of its lifecycle. The international reference standard is ISO 14067:2018, which builds on the ISO 14040/14044 lifecycle assessment framework but focuses specifically on the climate change impact category.

In practice, you're mapping emissions across a set of lifecycle stages. The standard ones look like this:

Raw material extraction and processing. The upstream emissions from mining, growing, or manufacturing the inputs to your product. For a concrete product, that's limestone quarrying, clinker production, cement grinding. For a packaged food product, it's agricultural inputs, fertiliser use, livestock emissions.

Manufacturing. Your factory floor. Electricity consumption, natural gas for process heat, fugitive emissions from refrigerants or industrial processes. This is usually where your Scope 1 and Scope 2 data already lives.

Transport and distribution. Getting raw materials to your factory, and finished product to your customer. Mode matters enormously here — sea freight versus air freight can change a product's footprint by an order of magnitude.

Use phase. What happens when the customer uses the product. For a building material, this might be negligible. For an appliance or a vehicle, it's often the dominant lifecycle stage.

End of life. Disposal, recycling, landfill. A product that ends up in landfill generating methane has a very different end-of-life footprint than one that's recycled into secondary material.

The system boundary you choose determines which stages you include. Cradle-to-gate covers raw materials through to your factory gate — the most common boundary for B2B products and EPDs. Cradle-to-grave covers the full lifecycle including use and disposal. Gate-to-gate covers only what happens in your facility.

Most companies start with cradle-to-gate because it's the boundary they can actually populate with reasonable data. We're honest about this: the further you get from your own operations, the worse the data gets.

The $50,000 Problem

Traditionally, getting a product carbon footprint meant commissioning an LCA consultant. A simplified study runs $5,000-$20,000. A full ISO 14040-compliant LCA for a moderately complex product — the kind that would hold up to critical review for an EPD — typically costs $30,000-$50,000 per product, sometimes more. Complex products with long supply chains can push past $100,000.

That pricing model works if you have one flagship product and a sustainability budget. It doesn't work if you have 50 SKUs and your customer wants footprint data for all of them.

It also creates a one-and-done problem. A consultant delivers a report. The numbers are valid for a period. Then your supplier changes, your energy mix shifts, your transport routes change, and the study is out of date. You either commission another study or keep citing stale numbers. Neither is great.

We built Carbonly's product-level carbon footprinting module because we kept hearing the same thing from manufacturers: "I need product carbon footprints, I can't afford $50K per product, and I need numbers that update when my inputs change." That's the problem statement.

Our approach uses lifecycle stages mapped to Scope 3 categories, with four system boundary options (cradle-to-gate, cradle-to-grave, gate-to-gate, and cradle-to-cradle). You populate each lifecycle stage with activity data — energy consumption, material quantities, transport distances and modes — and the system calculates emissions using Australian NGA factors and our material emissions library. Any stage contributing 10% or more of total product emissions gets automatically flagged as a hotspot, with AI-generated recommendations for reduction.

I should be clear about what this is and isn't. It's product-level carbon footprinting with lifecycle stages and hotspot analysis. It's not a full ISO 14040-compliant LCA tool like SimaPro or openLCA. Those tools model hundreds of impact categories, handle complex allocation procedures, and support the level of methodological detail required for third-party verified EPDs. Our module is designed for companies that need defensible product footprints without hiring an LCA specialist for every SKU — a middle ground between "we have no product-level data" and "we commissioned a $50K study."

For many companies responding to customer requests, preparing for CBAM compliance, or building an internal understanding of where product emissions concentrate, that middle ground is exactly what's needed.

Where the Emissions Actually Concentrate (It's Usually Not Where You Think)

The whole point of product-level footprinting is identifying hotspots. And the hotspots are almost always surprising.

Consider a food manufacturer producing a packaged sauce. They might assume the factory — gas-fired cooking, electricity for refrigeration, packaging lines — is where most emissions sit. That's what they see every day. But when you map the full cradle-to-gate footprint, the agricultural ingredients (tomatoes, spices, oils) and their associated fertiliser, irrigation, and land-use emissions often account for 60-80% of total product carbon. The factory itself might be 10-15%. The packaging perhaps 5-10%.

That changes your reduction strategy entirely. A $200,000 solar installation on the factory roof barely dents the product footprint if the agricultural Scope 3 is five times larger. You'd get more impact by switching to a lower-carbon oil supplier or sourcing from farms with different practices.

For construction materials, the pattern is different. Three materials — concrete, steel, and aluminium — account for over 70% of embodied carbon in a typical building, with raw material processing (the A1-A3 stages) dominating. A concrete manufacturer's hotspot is almost always the clinker production kiln. An aluminium fabricator's hotspot is smelting energy — which is why the state where smelting occurs matters enormously. An aluminium product smelted in Tasmania (grid factor 0.20 kg CO2-e/kWh) has a fundamentally different Scope 2 profile than one smelted in Victoria (0.78 kg CO2-e/kWh).

Transport mode is the other blind spot. We've seen products where switching from air freight to sea freight for a single input cut the transport-stage emissions by 95%. Companies that treat transport as a minor line item in their product footprint are often wrong — especially for products with lightweight, high-value inputs sourced internationally.

Australian Regulatory Drivers

Product carbon footprinting isn't explicitly mandated under AASB S2, but the standard's Scope 3 requirements push companies toward it. Category 1 (purchased goods and services) is typically the largest Scope 3 category for manufacturers and retailers. Calculating it properly means understanding the cradle-to-gate emissions of what you buy — which is, functionally, the product carbon footprint of your suppliers' outputs.

And it flows the other direction too. If your customer is calculating their Scope 3 Category 1, they need your product carbon footprint data. Spend-based estimates using industry-average emission factors carry 30-40% uncertainty. Product-specific data is what gets that uncertainty down to something defensible under assurance.

The NCC 2025 has introduced voluntary embodied carbon reporting for commercial buildings, with states adopting from May 2026. The NCC 2028 review may make embodied carbon limits mandatory. For building product manufacturers, this means product-level embodied carbon data will shift from "nice to have" to "required for specification." EPDs are the recognised format for communicating that data, and EPDs are built on product LCA.

For exporters, CBAM is the immediate pressure. Australian aluminium exports to the EU were worth US$4.7 billion globally in 2024. Even modest volumes to EU buyers require verified embedded emissions data under the definitive regime. Exporters who can't provide actual emissions data get hit with EU default values — which are typically set conservatively high, meaning you pay more than you should.

What You Actually Need to Get Started

Here's the practical reality of building product carbon footprint capability. It's less daunting than people assume, but it does require data you probably haven't collected before.

Start with your gate-to-gate. Your own operations are where you have the best data. Electricity bills, gas bills, process inputs, waste outputs. If you've already got a corporate emissions inventory through Carbonly or similar, most of this data exists. The shift is allocating it to specific products rather than reporting it as a site total.

Map your top 5 material inputs by mass or spend. You don't need emission factors for every input on day one. Start with the materials that make up the bulk of your product by weight or cost. For most manufacturers, 5-10 inputs account for 80%+ of the material footprint. Use supplier-specific data where available, industry-average factors where it isn't.

Choose your transport modes and distances. Where do your inputs come from? How do your products reach customers? Road, rail, sea, air, pipeline — each has a dramatically different emission factor per tonne-kilometre. You probably know this intuitively (air freight is expensive, and it's expensive because it uses a lot of fuel) but you might not have quantified it.

Accept that your first product footprint will have gaps. It will. Some lifecycle stages will use secondary data (industry averages) instead of primary data (your actual measurements). That's normal. The ISO 14067 standard explicitly allows secondary data where primary data isn't available. The point is to start, identify the hotspots, and progressively improve data quality for the stages that matter most.

Our module auto-recalculates whenever you update stage data, so improving one input — say, getting an actual emission factor from your steel supplier instead of using the industry average — instantly updates the product footprint. That feedback loop is what makes internal capability different from a one-off consultant study.

Honest Limitations

We build carbon accounting technology, and we think product-level footprinting should be accessible to more companies. But we won't pretend there aren't real limitations.

Secondary emission factors (industry averages, database values) introduce significant uncertainty — sometimes 30-50% for upstream agricultural or mining inputs. Your product footprint is only as good as the data going into each stage. For some lifecycle stages, good data simply doesn't exist yet, especially for end-of-life scenarios in the Australian context.

EPD generation still requires third-party verification under ISO 14025. Our module can give you the underlying product carbon footprint, but publishing an EPD through EPD Australasia means engaging a verified LCA practitioner for the critical review. We can feed the data into that process — but we're not a substitute for it.

For genuinely complex products with hundreds of components, multiple manufacturing sites, and intricate allocation challenges (co-products, recycled content, biogenic carbon), a specialist LCA tool and an experienced practitioner are still the right answer. We're honest about where simplified approaches break down. A semiconductor chip or a pharmaceutical product with 200+ precursors isn't something you should footprint with a simplified module.

And Scope 3 data from suppliers remains the biggest headache across the board. We've written about this at length — the practical challenges of getting supplier data don't disappear just because you're working at product level instead of corporate level.

The Middle Ground Is the Right Starting Point

Here's the argument we'd make. Waiting for perfect data before calculating product footprints means you'll have no product footprint data when your customer asks for it. And they're asking now — whether it's an EU importer needing CBAM compliance data, a construction project specifying materials by embodied carbon, or a retailer collecting supply chain footprints against SBTi targets.

A simplified product carbon footprint with clearly documented system boundaries, honest data quality indicators, and the ability to update as better data becomes available is far more useful than either (a) nothing, or (b) a $50,000 study that's accurate on the day it was written and increasingly wrong after that.

Commission the full ISO 14040 study for your flagship product or your EPD application. But for the other 49 SKUs in your catalogue? Build the internal capability to measure, track, and improve. That's where product-level footprinting with lifecycle stage mapping and hotspot analysis fits.

Start with one product. Your highest-volume product, or the one your biggest customer is asking about. Map the lifecycle stages, populate what you can, flag the hotspots, and improve from there. That first product footprint won't be perfect. It'll be useful — which is better.

Related reading: