Two ways to calculate a carbon footprint. Only one of them measures what actually happened.

There are two very different ways to arrive at a carbon emissions figure.

You can use activity data. That’s the actual kWh of electricity used, litres of fuel burned, miles travelled or tonnes of waste produced.

Or you can estimate emissions from how much money was spent.

They are not doing the same thing.

Activity-based reporting uses what actually happened. Spend-based reporting estimates emissions based on cost.

Both will give you a number. But one is built from real activity data and the other is an estimate.

That distinction matters.

What spend-based reporting actually measures

Spend-based carbon calculations are relatively straightforward.

Take the amount spent on a particular category of goods or services and multiply it by an emissions factor based on economic value.

Spend £X. Apply the relevant factor. Get X kg or tonnes of CO2e.

This isn’t some questionable methodology somebody invented on the back of an envelope. Spend-based calculation is recognised within GHG Protocol’s Scope 3 guidance.

But it is important to be clear about what the calculation is actually using.

It’s using money as the activity data.

GHG Protocol describes the spend-based method for purchased goods and services as taking the economic value of the purchase and multiplying it by an appropriate environmentally-extended input-output emissions factor.

That’s an estimate.

And GHG Protocol’s own guidance makes an important distinction between primary and secondary data.

Primary data comes from specific activities within a company’s value chain. Secondary data includes industry averages, financial data, proxy data and other generic information.

Spend-based calculations sit in that second group.

[Read the GHG Protocol Scope 3 Calculation Guidance]

Here’s the problem with using spend to measure change

Prices move.

That’s normal.

Inflation pushes prices up. Suppliers put prices up. Contracts are renegotiated. Discounts disappear. Energy prices rise and fall. Businesses change supplier.

None of those things necessarily tells us whether the underlying activity changed.

Imagine a business uses exactly the same amount of a product two years running.

Year one costs £10,000.

Year two costs £12,000.

Nothing changed except the price.

Put those figures through a spend-based carbon calculation and the emissions estimate can change too.

Has the business actually increased its carbon emissions?

You don’t know.

That’s the problem.

The same thing can happen in reverse. Negotiate a better price and a spend-based calculation can fall without the underlying physical activity changing at all.

That makes spend-based data useful for estimating emissions where better information isn’t available.

What it doesn’t make it particularly good at is telling you whether the physical activity responsible for emissions has actually changed.

GHG Protocol itself warns that financial spend and emissions may not correlate well. Some high-value activities can have relatively low emissions. Some lower-value activities can have a significant GHG impact.

That’s quite an important limitation when the number you’re producing is supposed to help you understand carbon.

[Read the GHG Protocol Scope 3 FAQs]

Activity data gives you a different picture

Now take the money out of the calculation.

Instead of asking what the electricity cost, use the kWh.

Instead of what was spent at the fuel station, use the litres of fuel.

Instead of the cost of business travel, use the actual distance travelled where that data is available.

Instead of the invoice value for waste collection, use the actual weight and waste type.

Now when the underlying activity changes, the data changes with it.

If electricity consumption falls from 50,000 kWh to 40,000 kWh, you’ve got something meaningful to compare.

If diesel consumption falls from 10,000 litres to 8,000 litres, you can measure that change.

And if consumption doesn’t change, the activity data doesn’t suddenly tell you that it did because somebody put their prices up.

That’s why I believe activity-based data gives businesses a much stronger foundation for carbon reporting.

It is based on what happened.

And if you’re serious about reduction, that matters

A carbon reduction plan should ultimately answer a fairly simple question:

Are our emissions actually coming down?

To answer that properly, you need to be able to see what changed.

Was less electricity used?

Did fuel consumption fall?

Did mileage reduce?

Did waste decrease?

Did a change in equipment actually reduce energy consumption?

Those are operational questions.

A change in spend cannot reliably answer them because spend is affected by far more than physical consumption.

GHG Protocol makes essentially the same distinction when discussing data quality.

Its Scope 3 guidance says primary data provides better representation of a company’s specific value-chain activities. It also says primary data allows companies to track operational changes resulting from emissions-reduction actions and better track progress towards GHG reduction targets.

That’s significant.

If the purpose of carbon reporting is eventually to reduce carbon, surely the data capable of showing an operational change should be the data we work towards using wherever it’s available.

Spend-based has a use. I don’t believe it should be the destination.

There is a perfectly practical reason spend-based reporting exists.

Sometimes the activity data simply isn’t available.

A large company trying to understand thousands of purchases across a complicated supply chain isn’t necessarily going to have actual physical data for everything it buys.

Spend data is already sitting in the accounts. It’s accessible. It’s relatively easy to categorise. And it allows an organisation to produce an estimate where otherwise it might have nothing.

That’s useful.

But there’s an important difference between saying:

“We don’t have better data, so we’ll estimate this.”

and saying:

“This estimate is as good as measuring the activity itself.”

It isn’t.

Even GHG Protocol’s guidance says companies should generally collect high-quality primary data for high-priority activities, and that primary data should be used for Scope 3 activities targeted for GHG reductions.

So perhaps the question isn’t whether spend-based reporting should exist.

It is:

Where real activity data is available, why would we choose not to use it?

The data already exists in a lot of businesses

This is the bit I think gets missed.

Activity-based carbon reporting can sound as though businesses need to start collecting an entirely new world of information.

Often, they don’t.

Businesses already receive electricity bills showing kWh.

They already have gas bills.

They already buy fuel by the litre.

They already record business mileage.

They already receive waste documentation.

They already have refrigeration and service records.

The activity happened and, in many cases, somebody recorded it because the business needed that information for completely ordinary reasons.

The problem has been turning those records into usable carbon data without creating another enormous administrative job.

That’s a very different problem from the data not existing.

Better supplier data means better Scope 3 data too

This becomes particularly important when carbon data moves through a supply chain.

GHG Protocol distinguishes between primary data from specific value-chain activities and secondary data such as industry averages and financial information.

It says primary data can come directly from suppliers and can include activity data or emissions figures calculated from a supplier’s specific activities.

It also recognises the practical problem: where supplier-specific data isn’t available, companies have to use secondary data instead.

So if a supplier cannot provide actual emissions information, somebody further up the chain may have to estimate that part of their Scope 3.

Again, the estimate serves a purpose.

But improving the quality of the data at supplier level improves the information available further up the chain too.

This is one of the reasons I think activity-based reporting will become increasingly important.

Not because every estimate is suddenly going to be banned.

Because better primary data is becoming increasingly possible.

Where I think carbon reporting is heading

This part is my opinion.

I think we’ll look back at widespread spend-based carbon estimates as a necessary stage in the development of carbon reporting.

They gave organisations a practical way to start putting numbers around emissions when primary data was difficult, expensive or sometimes impossible to obtain.

But technology is changing that.

If activity data can be extracted from the records businesses already hold without somebody manually spending hours entering it into spreadsheets, one of the biggest practical arguments for relying on spend starts to weaken.

And if the purpose of all this is genuinely to reduce emissions, rather than simply produce a carbon number, I think the direction makes sense.

More primary data. More activity-based reporting. Fewer estimates where actual information is available.

GHG Protocol itself is currently revising its Scope 3 Standard. The March 2026 Phase 1 Progress Update contains proposals around greater data granularity and transparency. These are proposals, not final requirements, and the revision process is ongoing.

So I’m not going to pretend I know exactly what the final standard will say.

But the broader question isn’t going away.

What is the carbon number actually based on?

[Read the GHG Protocol Scope 3 Standard Revisions Phase 1 Progress Update]

Measurement before reduction

Spend-based reporting can give you an estimate.

Sometimes that’s exactly what you need.

But if you have the actual activity data, you have something better.

You can see what was used.

You can calculate the emissions associated with it.

You can compare one reporting period with another.

And when you make a change intended to reduce emissions, you can see whether the underlying activity actually changed.

For me, that’s the difference.

If we’re serious about reducing carbon emissions, we need to get as close as we reasonably can to measuring what actually happened.

Because it’s very difficult to prove you’ve reduced something you weren’t really measuring in the first place.