Light the Spark

Annie Freeman, Public Speaker | Engineer at Xero | Green Software Advocate

The journey began with a simple realization: while climate change might seem like someone else’s problem, it requires every industry to transform how they operate. As someone passionate about making an impact, I wanted to connect my full-time software engineering role with environmental action.

When you work in a large software company, adding another initiative to an already packed roadmap can feel overwhelming. This was exactly the challenge we faced when considering how to integrate sustainable computing practices into our ways of working. Without a direct feature benefit to customers, these initiatives often risk falling to the bottom of priority lists. However, with 99% of our employees expressing deep concern about sustainability in our company survey, we knew we had to find a way forward.

We started with a hackathon project that would eventually transform our approach to software sustainability. We focused on what matters most: data.

Cloud providers’ emissions data, while improving, often lacks the granularity needed to drive real change. We needed to make this personal—to show engineers exactly how the software components they own impacted our carbon footprint. To develop a proof of concept, we worked with Climatiq to get emissions factors data, and leveraged internal AWS billing data to generate granular data. This would eventually evolve into Footprint, our internal measurement tool.

The impact of having concrete data was immediate and powerful. A perfect example came when one of our team members examined their emissions data and discovered that their choice of using AWS CodeBuild for remote code execution was having an outsized impact on carbon footprint and costs. This revelation prompted them to redesign their architecture around Kubernetes, significantly improving both environmental impact and operational efficiency.

Building on this momentum, we established a community of practice that grew to over 70 members within a few months. We ran educational sessions, shared insights about data center efficiency, and encouraged participation in the Linux Foundation’s Green Software Practitioner training. Our internally built Footprint tool became increasingly sophisticated, providing teams with component-level emissions data that spoke directly to their work, including a breakdown of which cloud services contributed to the emissions output of each component.

The current phase of our journey focuses on normalization—making carbon emissions just another “boring” metric that teams routinely consider alongside performance, reliability, and cost. Engineers now roll emissions data into their technical decisions and sprint reviews.

Looking ahead, we are partnering with our sustainability team to create a comprehensive engagement plan. Our future plans include developing a self-assessment tool that evaluates software sustainability maturity, organizing competitions and hackathons, and integrating green software practices into our engineering standards.

The key lesson from our journey? Sometimes, the most effective approach to transformation isn’t a dramatic overhaul but rather a series of small, data-driven steps that gradually reshape how we think about and build software. By making sustainability tangible and personal to each engineer’s daily work, we’re seeing real change take root in our engineering culture.