Sustainability teams are often measured by how much they own. Tamara Somers, Global Head of Sustainability and Impact at Xero, sees success differently. For her, the goal is not to build a sustainability function that permanently manages every ESG topic, but to create the structures, knowledge and processes that allow the wider organisation to take ownership. This philosophy runs through Xero’s approach to carbon accounting, Scope 3, climate targets, regulation and artificial intelligence, and it points to a broader shift in sustainability management: away from sustainability as a specialist reporting exercise and toward sustainability as part of how the business itself operates.
1. Build the Capability, Then Give It Away
When Tamara joined Xero, one of the first challenges was deciding what the sustainability team should focus on. The potential agenda was almost unlimited, which made prioritisation essential. As she explains, sustainability professionals can easily end up trying to do everything precisely because they care about the work. Xero introduced a deliberately simple mechanism for ideas that mattered but were not yet priorities: a shared file called the “not yet bucket”. The principle behind it was equally simple: do a few things well rather than trying to do everything poorly.
That focus also shapes Tamara’s view of what a mature sustainability function should ultimately become.
“I often describe my vision of a sustainability team as being a sort of incubator for the business. You don’t want all the work to stay with the sustainability team because then it’s kind of off to the side. You want once people understand it and know how to do it, you really want the business, you know, the line areas that are responsible for executing it to take ownership.”
At Xero, that is already happening. Diversity and inclusion sits within the people function, supplier risk management and supplier carbon work sit within procurement, and carbon accounting and ESG data management have moved into finance because that is where Tamara believes they need to live if they are to become core to how the organisation works. That shift from ownership to orchestration is important for sustainability technology too. A platform should not simply create a more sophisticated ESG silo; it should help sustainability information move into the functions that actually control expenditure, suppliers, operational decisions and financial planning. For Footprint Intelligence, this points to a broader role for sustainability software: not simply calculating sustainability metrics, but creating the infrastructure through which different parts of the organisation can work with them.
2. From Carbon Accounting to Business Decisions
This shift is especially visible in Xero’s carbon accounting. Around 99% of its emissions sit in Scope 3, which means emissions management is inseparable from supplier data, procurement and the quality of information coming from outside the sustainability team. Scope 3 is therefore not simply a calculation problem, but a coordination problem. Companies need supplier information, financial and activity data, emission factors, assumptions, evidence and increasingly assurance-ready documentation, and all of that has to come together before the organisation can make credible decisions about where to act.
Xero has moved carbon accounting into its finance function and applied finance’s reporting discipline to it. Tamara believes the result is more rigorous, and not simply because of the financial expertise. Accounting brings stronger working papers, documented assumptions and a clearer basis of preparation. Even with software in place, though, a substantial amount of manual work remains. That is why Xero is now looking at how AI, particularly agentic AI, can take on more of it, so sustainability professionals spend less time reconstructing information.
“No one really got into this job to do that.”
This is a useful benchmark for sustainability software. Automation should not merely make reporting faster; it should reduce repetitive data work so sustainability professionals can focus on suppliers, reduction measures, targets and business decisions.
The same need for discipline appears in climate target setting. In Australia, forward-looking statements require a reasonable basis, which means climate targets need to be supported by a credible pathway rather than ambition alone.
“We have to show them how we’re going to get there for them to approve it.”
For a fast-growing software business, that is not straightforward. Xero’s emissions pathway depends on assumptions around spending, headcount, marketing, digital advertising and the changing technology stack. Generative AI has become a particularly important new variable because it was not material when parts of Xero’s earlier emissions models were built. That uncertainty has already reached board level, with Tamara being asked whether Xero can still be confident in meeting its targets given the impact of AI. Her answer was deliberately cautious: the company does not yet have enough data to know.
This is arguably what credible sustainability management increasingly looks like: not claiming greater certainty than the data allows but building models that are explicit about assumptions and can evolve as the business changes. The same pragmatism applies to decarbonisation. Xero’s total footprint is relatively small compared with major industrial emitters, while business travel represents around 20% of its emissions. In Australia and New Zealand, however, aviation is difficult to avoid. Xero therefore focuses on interventions it can realistically influence, including sustainable aviation fuel through a book-and-claim system with partner airlines. The broader lesson is that a carbon footprint tells a company where emissions are; a sustainability management system needs to help determine what can realistically be changed next.
3. AI, Reporting and the Value Chain
Few topics illustrate the changing sustainability landscape better than AI. On the one side, AI could remove significant manual work from sustainability teams. On the other, its energy, water and emissions footprint is creating an entirely new measurement challenge. Xero is already seeing large differences in the environmental data available from technology suppliers. Some providers can supply detailed information about energy, water and emissions, while others, as Tamara puts it more bluntly, “basically tell you to go away when you ask for it.” Where direct data is unavailable, Xero has to rely on modelling and academic research. The underlying problem is that there is still no set standard for measuring the environmental impact of AI, and estimates can vary significantly depending on the model, data centre and methodology being used.
Xero’s response has been to first establish a credible baseline. One conclusion from that work is particularly relevant for companies rapidly increasing their use of AI:
“The model you choose is more important than the provider in terms of the environmental impact. So I think the first message for people is: choose the right tool for the job.”
Tamara gives a deliberately everyday example: “Don’t use Claude to find out where you want to go for lunch when a Google search will do that.” Behind the humour is a serious management point. More powerful models can come with larger inference footprints, so organisations need to consider whether the sophistication of the model is justified by the task. This is also where environmental and financial efficiency begin to reinforce each other. Tamara argues that sustainability teams are more likely to succeed when reducing environmental impact also reduces cost, rather than relying on either argument alone. That principle goes well beyond AI: sustainability initiatives are often easier to scale when the environmental rationale is connected to lower cost, reduced risk, greater resilience or better operational performance.
Tamara takes a similarly pragmatic view of mandatory sustainability disclosure. Reporting is often criticised for absorbing resources that could otherwise go into implementation, but she challenges the idea that reporting and “real work” are opposites. In her view, meaningful reporting depends on there being substantive work underneath it in the first place. Australia’s mandatory climate disclosure regime has forced companies to build stronger governance, processes, documentation and internal consultation mechanisms. This creates substantial work, particularly in the first reporting cycle, but it can also bring sustainability issues onto the board agenda and create pressure for action. Tamara describes herself as “a bit of an optimist about mandatory disclosure as an engine for getting the work done that needs to be done”, while also recognising the significant compliance burden it can create, particularly for smaller teams.
That burden becomes even more important once sustainability requirements move through the supply chain. As Scope 3 reporting and assurance requirements increase, large companies need stronger evidence from their value chains, which means asking suppliers for more data.
“That means that I have to go and ask everyone in my supply chain for their emissions data, and if you’re a small business that’s a really difficult thing to be able to provide.”
This is where regulation aimed formally at large organisations can become a practical compliance requirement for much smaller companies. Even where legal thresholds exclude them directly, supply-chain requirements can still pull them in. Tamara describes the result as a “web of compliance” in which smaller suppliers can become subject to demands that technically originate with larger organisations.
That has an important implication for sustainability technology. Supplier engagement cannot remain an annual exchange of emails and spreadsheets if companies are increasingly expected to produce robust Scope 3 information and evidence. They need structured ways to request supplier data, assess its quality, combine it with other calculation methods and make it usable for both reporting and decision-making.
From Sustainability Function to Sustainability Infrastructure
Tamara’s “incubator” analogy provides a useful way to think about where sustainability management is heading. The objective is not to build ever-larger sustainability teams that manually own every process, but to establish the governance, data, tools and capabilities that allow sustainability to become integrated into finance, procurement, operations and management.
That also changes what sustainability software needs to do. Calculating emissions remains important and reporting remains important. But increasingly, those are only parts of a broader system: one that needs to collect and validate data, engage suppliers, support credible target pathways, connect reduction measures with financial implications and distribute responsibility across the business.
For Footprint Intelligence, this points toward a broader role for sustainability software: not another specialist system sitting beside the business, but infrastructure connecting carbon data, suppliers, targets, measures, financial implications and the responsibilities across it. The strongest sustainability systems may ultimately be the ones that make the sustainability team less central to day-to-day execution, precisely because it has become more central to how the company operates.
