A net zero plan for manufacturing cannot succeed on ambition alone. Boards may want meaningful progress on decarbonisation, while CFOs need clear evidence that each step makes financial sense. The challenge is not choosing between those goals, but using better data and modelling to build a plan that delivers both.
I was at a manufacturing industry event recently where a panel discussion on decarbonisation turned into what can only be described as a polite argument.
On one side, the sustainability advocates. “This is an existential imperative. We have to act now. The planet doesn’t care about your payback period.”
On the other, the finance people. “We’re running a business. We can’t invest millions in equipment changes that won’t pay for themselves for a decade. Our shareholders won’t accept it.”
Both sides were passionate. Both sides had a point. And both sides were talking past each other because they were using completely different frameworks to evaluate the same decisions.
This is the conversation happening in manufacturing boardrooms across the UK right now. And it’s stuck.
The gap between ambition and action
Most UK manufacturers have made some form of net zero commitment. It’s in the annual report. It’s on the website. The board has discussed it. The intent is genuine.
But when you look at what’s actually happened beyond the statement, in many cases the answer is: not much. Some LED lighting. Maybe some solar panels. A switch to a green energy tariff. The easy wins have been taken.
The hard stuff (process changes, equipment replacement, supply chain decarbonisation, heat recovery, electrification of gas-fired processes) remains in the “we should look at that” category. Not because anyone’s opposed to it, but because the economics feel uncertain, the options feel overwhelming, and nobody’s quite sure where to start.
The sustainability team (if there is one) talks in tonnes of carbon. The finance team talks in pounds and payback periods. The operations team talks in production risk and downtime. And the CEO is caught in the middle, trying to reconcile three legitimate but seemingly incompatible perspectives.
Why most decarbonisation roadmaps gather dust
I’ve reviewed more net zero roadmaps than I’d like to admit. Most of them share a common flaw: they’re long on aspiration and short on financial modelling.
They’ll tell you that you need to reduce Scope 1 emissions by 40% by 2030. They’ll list the interventions required: heat pumps, electrification, process changes, renewable energy procurement. They might even give you a timeline.
What they rarely give you is a credible financial model. What does each intervention actually cost? What’s the realistic payback, given your specific energy consumption profile? Which interventions interact with each other? Does investing in heat recovery change the economics of electrification? What funding is available, and how does it affect the numbers? What’s the impact on cash flow, year by year?
Without that modelling, the roadmap is an aspiration, not a plan. And aspirations don’t survive contact with a quarterly board meeting where the CFO is looking at capital allocation decisions.
This is a data problem with a data solution
Here’s what I find genuinely exciting about this space as a data scientist: the decarbonisation challenge in manufacturing is fundamentally a modelling problem. And modelling problems are what AI does extraordinarily well.
Every manufacturer has energy consumption data. They have production output data. They have cost data. They have (or can get) carbon conversion factors. They have capital expenditure forecasts. They have information about available grants, tax incentives, and green finance products.
The challenge isn’t collecting this data. It’s connecting it into a model that lets you ask meaningful questions. Questions like: if we invest £200k in heat recovery this year, what’s the impact on our energy bill over five years, taking into account projected energy price trends? If we electrify our two highest-emission processes, what’s the carbon reduction, the capital cost, and the net financial impact after available grants? If we do nothing for three years and then try to catch up, what does the cost curve look like compared to starting now?
These aren’t hypothetical questions. They’re the exact questions that boards need to answer in order to move from commitment to action. And they’re the questions that a well-built AI model can answer with specificity, sensitivity analysis, and scenario planning that no spreadsheet can match.
Finding the right sequence
The biggest insight that modelling consistently delivers is this: the sequence matters enormously.
Not all decarbonisation interventions are equal. Some are high-impact and low-cost. Some are low-impact and high-cost. Some only make financial sense after you’ve done something else first. The right sequence of interventions can be dramatically cheaper than the wrong one, sometimes by a factor of two or three.
But you can’t see the right sequence without the model. And you can’t build the model without connecting the data.
This is where the CFO and the sustainability advocate finally end up in the same conversation. Because when you can show the board a modelled pathway that reduces emissions by 35% over five years, with a positive cumulative NPV, funded through a combination of operational savings, available grants, and green finance, that’s a plan the whole board can get behind.
Not because anyone compromised. Because the data showed them a path that satisfied both objectives simultaneously.
Decarbonisation doesn’t have to be a leap of faith
The manufacturers who are making real progress on net zero aren’t the ones with the biggest budgets or the boldest commitments. They’re the ones who’ve done the modelling. They know their numbers. They’ve identified the highest-impact, lowest-cost moves. And they’ve built a sequenced plan that makes financial sense at every stage.
If your net zero commitment still feels more like an aspiration than a plan, the missing ingredient probably isn’t willpower. It’s data.
That’s a gap we can help close.
Author: Peter Appleby, Co-Founder & Chief Data Scientist.
Follow our LinkedIn company page to stay up to date with all our new blogs!
