Sustainability is not a department

Most companies in tech approach sustainability the same way. A dedicated team is appointed, a consultant is briefed, a deck is produced, and the rest of the organisation is invited to come along. Two years later, the targets are still on the website, and most of the actions that would have moved them are still sitting in someone's backlog.

The problem is that sustainability questions, when you actually look at them, sit at the operational core of how a business runs. They are not adjacent to it. And a plan built next to the work, rather than inside it, has a hard time turning into anything real.

We saw this clearly when we ran our own carbon assessment at efficy Group earlier this year. The headline numbers told a familiar story for a European software group. People transportation accounts for 37% of our footprint. The downstream use of our software and services accounts for 19%. IT and digital services, dominated by cloud hosting, account for 15%. Office consumption, miscellaneous purchases and other categories fill out the rest.

The map made one thing obvious: the decisions that actually shape these numbers are made in three completely different parts of the company. So that's where we went to think about them, in three separate workshops, with the people who already hold the levers. Here is what came out of those conversations.

Digital: compute hasn't been free for a while

For most of the last twenty years, software engineering has treated compute as essentially free. You needed more, you provisioned more. The cost was somewhere else, in the data centre, on the electricity grid, in the atmosphere. That assumption is breaking, and it is breaking faster now that AI agents run continuously across customer data, calling other systems and iterating without waiting for a user to click.

The implication for how we build is that energy can no longer be treated as something you optimise at the end. By the time a product is in production, the architectural choices that shape its energy profile have already been made. Hosting region. Compute density. Storage patterns. The volume and frequency of background processes. A system designed without that constraint is cheaper to build and more expensive to run, year after year, until someone has to redesign it.

This is what we mean when we talk about green architecture, and it's why the work on this sits with Product and Tech, not with a sustainability function. The concrete priorities that came out of the workshop reflect that. Identifying and decommissioning idle cloud resources, which has both an environmental and an economic return. Optimising the geographical choice of hosting regions based on the carbon intensity of their energy mix. Extending the lifespan of IT equipment through refurbished sourcing and longer renewal cycles. And beginning a deeper analysis of how our products themselves consume energy, with the goal of making eco-design a real input into the way we build, not an afterthought.

Energy becomes a real variable in design reviews, infrastructure decisions, and the trade-offs we make between shipping fast and shipping well. That's a job for the people who build the systems.

Mobility: a calendar problem, not a poster problem

The second category that came out of our assessment is people transportation. It is also the one most companies handle the same way, with a charter, a poster in the kitchen, and a paragraph in the careers page.

The reality of mobility, when you sit with the data, is that it is made of thousands of individual decisions. The car each colleague drives. Whether a meeting in Stockholm justifies a flight or fits on a train. Whether there is a real alternative to the car for the commute. None of these decisions get changed by a top-down statement. They get changed by making the alternatives genuinely available and the defaults genuinely different.

That is why this work sits with People, not with a sustainability office. The priorities from the workshop are practical in the same way. Tightening the criteria in our car policy as we continue moving the fleet to electric, with weight and efficiency thresholds that ensure a real environmental gain rather than a symbolic one. Establishing rail as the default for journeys where it works, instead of a virtuous exception. Backing the colleagues who already cycle, take public transport, or share rides, with something more concrete than a thank-you in the all-hands.

Changing how a company moves means changing what shows up in everyone's calendar, not only what shows up on the careers page.

Procurement: you can't decarbonise what you can't measure

The third category we looked at covers everything we buy and the buildings we sit in. It is also the area where the gap between intention and reality is widest in most sustainability plans, because the data is usually the weakest.

A meaningful share of our emissions sits in services we purchase from other companies, and in offices where the energy bill is paid by the landlord. The carbon footprint of those activities is, in many companies, estimated from industry averages rather than measured. That is fine for a first assessment. It is not fine as a basis for action.

This is the work the Finance team picked up. Strengthening the data we collect on what we actually buy, so it can flow from accounting in a usable form. Surveying our top suppliers on their real emissions, so we move beyond industry averages. Defining a sustainable procurement policy with weighted carbon criteria in our tender processes. And establishing green clauses in our office leases, so the buildings we occupy are part of the conversation, not a black box.

A plan you cannot measure is a plan you cannot manage. The procurement workshop was, in many ways, a measurement workshop first.

Sustainability is not a department

What ties these three conversations together is a shared posture rather than a shared methodology. We chose to treat sustainability the way we treat the rest of the business: with data we trust, accountable people, and budget engaged against the changes we want to make. Targets will follow that work. They won't lead it.

That is not the fastest way to build a sustainability plan. It is, in our experience, the only way to build one that survives contact with the operational reality of a company. The deck is less neat at the end, but the actions are real.

If sustainability isn't sitting in the calendars of the people who run the business, it isn't happening.

Take a look at our sustainability approach