Op-ed by Oskar Klingberg, co-founder, Bardo.
Every chief sustainability officer I meet has a target. A net-zero date, a science-based reduction curve, a number they have committed to publicly and that the board now expects them to deliver. What far fewer of them have is a way to see which of their company’s thousands of daily decisions actually moves that number.
That gap, between the ambition set in the boardroom and the action that has to happen everywhere else, is the defining frustration of the role. For most of the last decade it has been treated as a problem of will, culture or budget. It is really a problem of data.
Consider how corporate emissions have traditionally been measured, especially the indirect Scope 3 emissions that make up the overwhelming majority of most companies’ footprints, often around 95% The standard method is spend-based: take what you spent in a category, multiply it by an average emissions factor, report the result.
For a first report, it may have seemed like a sensible place to start. But to actually inform a decision, it is close to useless.
A spend-based model cannot tell two products apart if they cost the same. It cannot distinguish a supplier running on renewable power from one burning coal. And it has a perverse quirk that anyone who has lived through the last few years will recognise: when prices rise, reported emissions rise with them, even if you bought exactly the same things, or even greener ones.
The logic is simple: if you cannot connect a decision to its consequences, you cannot steer by it.
Pushing against fog
A real example makes the point. A MacBook Air, by Apple’s own reporting, carries a carbon footprint of around 150 kg of CO2e. HP’s own figure for its 17-inch laptop is roughly double that, a mean of 310 kg. The catch is that the MacBook is the more expensive machine.
A spend-based model, which infers emissions from what you paid, would therefore hand the pricier but lower-carbon laptop the bigger number. It does not merely fail to tell two products apart. It can rank them exactly backwards.
Multiply that blindness across everything a company develops, sources and sells, and you see why so much sustainability effort can feel like pushing against fog. The choices that move the footprint are made constantly, and the data has rarely been good enough to show which ones counted.
Faced with that, the obvious move is to measure at the product level instead. The data that comes from doing it properly is genuinely good. The problem has always been getting it: until recently, product-level accuracy meant heroic manual effort, detailed lifecycle assessments built by hand, product by product, each one a project in itself.
Accurate, but so slow and expensive that it could only ever cover a handful of flagship products, never the full, shifting flow of everything a company actually buys. You could have accuracy or you could have coverage. Not both.
A practical shift
This is now changing, and the shift is more practical than most people expect. The raw material is data a company already keeps: the invoices, ERP entries and purchase records that describe, in commercial detail, exactly what was bought and from whom. The hard part has always been turning that mountain of unstructured documentation into real emissions, product by product, at a scale no team could handle manually.
That is now work for agentic AI. The software gathers the documentation behind each purchased product or service and reconstructs what it actually emitted: a flight built up from the aircraft type and its fuel burn, the departure and destination airports and the distance between them, rather than from a spend average; a single pen from the materials it is made of and where those materials were produced.
Human specialists review the work, handle the corner cases, and feed their corrections back so the system keeps getting better. The output is the real number, traceable to a source document and defensible in an audit.
The effect on a sustainability leader’s work is significant. Emissions stop being a once-a-year estimate assembled long after the fact and start to be accounted for the way money is: from the actual transactions, and the real products and suppliers behind them. A greener choice finally shows up as a lower number, instead of disappearing into an average.
For a CSO, this is the difference between describing the problem and being able to act on it. It also changes the company’s relationship with its suppliers. Once emissions intensity is visible at the level of individual products and suppliers, it becomes a factor in decisions alongside price and quality.
Suppliers notice. They begin to measure and reduce, because their customers are now asking, and now able to tell the difference. A low footprint becomes a competitive advantage worth investing in.
That is the quiet mechanism that may end up driving more decarbonisation than any single regulation: demand, flowing up the supply chain, rewarding the companies that emit less. It only works if the buyers can see clearly. For most of the history of corporate sustainability, they could not.
Refocusing efforts
What this changes, in the end, is where a sustainability leader’s efforts go. For years the role has been consumed by producing and defending data, turning messy records into a number that survives scrutiny.
Hand that work to a system that does it continuously, and the team is free to do what the data was always meant to enable: cut emissions, by redesigning products and choosing better suppliers. And it is a timely shift, because emissions are fast becoming something a company’s own customers weigh in their decisions.
Those that can see their footprint clearly, and act on it, will be the ones that stay competitive as that expectation hardens.
Learn more about Bardo at www.bardo-technology.com.
