ESG reporting is a cost centre. It does not have to be.

Ask a sustainability lead what their annual reporting costs and you will get a number for the consultant engagement. Ask what it returns and the honest answer is a document, filed, that nobody in operations ever opens.

This is a strange outcome, because the underlying data — metered energy consumption, by site, by period — is the same data that tells you where your money is going. One organisation buys it twice and uses it once.

What the annual exercise actually consists of

The expensive part of a carbon report is rarely the emissions factor arithmetic. That part is a multiplication. The expensive part is assembly, and it is worth spelling out in full, because it is where the invoice comes from.

• Finding the bills. Utility statements across sites, suppliers and formats, some in a shared drive, some in an inbox, some only obtainable by asking the landlord.
• Reconciling entity names. The same site appears as three different strings in the lease register, the finance system and the utility account, and a human has to decide they are the same building.
• Resolving boundaries. Which sites are in the reporting entity for which months, what happens to one that changed tenancy in August, and how a shared landlord supply is apportioned between occupiers.
• Chasing gaps. The facility whose meter readings stopped in March, the estimated bill that was never trued up, the quarter where the supplier changed and the account number with it.
• Deciding what to do about the gaps. This is the step that quietly determines the answer. An estimate is made, it goes into a spreadsheet cell that looks identical to a measured one, and by the following year nobody can tell which cells were which.
• Building an audit trail. Enough evidence that somebody outside the organisation could follow the arithmetic back to a source document.

That work is repeated annually, largely by hand, and its output is a spreadsheet that is authoritative for about six weeks. The following year it is repeated from scratch, because nothing structural was built. The cost is not high because carbon accounting is difficult; it is high because the exercise begins each year from documents rather than from a system.

Where metered building energy sits in the scopes

The GHG Protocol Corporate Standard is the framework nearly every other regime builds on, and it covers seven greenhouse gases: carbon dioxide, methane, nitrous oxide, hydrofluorocarbons, perfluorocarbons, sulphur hexafluoride and nitrogen trifluoride. It divides emissions into three scopes.

For a commercial property portfolio, the mass of the reportable figure sits in Scope 2, and the GHG Protocol's Scope 2 Guidance exists specifically to standardise how organisations measure emissions from purchased electricity, steam, heat and cooling. That is the part a metering platform can put on a defensible footing, because it is the part that is already instrumented. Scope 3 is a different discipline with different inputs — the Corporate Value Chain (Scope 3) Standard defines fifteen categories of upstream and downstream activity, and a building meter speaks to almost none of them.

Location-based and market-based, and why the difference is commercial

Scope 2 has two accounting methods, and confusing them is one of the more expensive errors in corporate reporting because it is invisible until somebody checks. The Scope 2 Guidance puts the distinction plainly: "the market-based method reflects emissions from electricity that companies have purposefully chosen (or their lack of choice), while the location-based method reflects the average emissions intensity of grids on which energy consumption occurs."

The location-based figure is a statement about where you consume. The market-based figure is a statement about what you bought. They answer different questions and they move for different reasons — and the guidance requires companies with operations in markets offering contractual instruments to report both, labelled by method, which it terms "dual reporting".

Why this matters commercially, rather than merely technically:

• A purchasing decision can move one number and not the other. Signing a renewable supply contract can change the market-based figure substantially while the location-based figure, which reflects the physical grid, barely moves. An organisation that reports only the market-based number can appear to be decarbonising while its actual demand on a fossil-heavy grid is unchanged.
• An efficiency measure moves both. A kilowatt-hour you do not consume is absent from both methods. This is the honest lever, and it is the one that also reduces the bill.
• The market-based method has an evidence bar. The guidance sets out eight Scope 2 Quality Criteria that contractual instruments must meet to be a reliable data source for that method. Contracts that do not meet them are not a shortcut.
• Both methods still require the denominator. Whichever factor you apply, you apply it to kilowatt-hours. If the consumption series underneath is estimated, gappy or double-counted, the choice of method is beside the point.

That last point is the one we care about. Emissions factors are published and maintained by others — the US EPA's GHG Emission Factors Hub, for instance, is updated annually and keeps prior versions archived, which is exactly the property a reproducible calculation needs. The variable nobody else can supply for you is your own consumption.

The direction of travel is assurance

Sustainability reporting is moving from disclosure towards audited disclosure, and the institutional scaffolding for that is now largely in place.

The ISSB's standards set the disclosure baseline in a growing number of jurisdictions: IFRS S1 covers general requirements, including disclosure of an entity's performance against any targets it has set or is required to meet, and IFRS S2 covers climate specifically. In the European Union, the Commission states that EU law "requires companies above a certain size to disclose information on what they see as the risks and opportunities arising from social and environmental issues", with reporting under the European Sustainability Reporting Standards developed by EFRAG — the Commission's own page is the place to check current scope, which continues to be revised.

In our own region, the Securities Commission Malaysia's National Sustainability Reporting Framework adopts IFRS S1 and IFRS S2 as the baseline standards for Malaysian companies, applies them through "a phased and developmental approach", takes a climate-first approach with transition reliefs for harder disclosures such as Scope 3, and records that work is under way towards a Sustainability Assurance Framework. We are deliberately not restating scope thresholds or dates here: they are being revised in several jurisdictions at once, and the regulator's page will be right when this article is not.

On the assurance side, the IAASB has issued ISSA 5000, a standard for sustainability assurance engagements that is "suitable across all sustainability topics, information disclosed about those topics, and reporting frameworks" and that addresses both limited assurance and reasonable assurance. Separately, CDP runs the long-standing voluntary disclosure system that many organisations already report into, describing itself as connecting reporting organisations, public authorities and financial institutions through shared data infrastructure.

What limited and reasonable assurance ask of your data

The two levels differ in how much evidence the practitioner gathers, and that difference lands directly on the person who has to supply it.

Under limited assurance, the practitioner performs procedures — typically enquiry and analytical procedures — sufficient to conclude that nothing has come to their attention suggesting the information is materially misstated. It is expressed negatively. In practice the questions are about whether a process exists and appears coherent: where does this figure come from, what did you do about gaps, why does the third quarter step up.

Under reasonable assurance, the practitioner gathers sufficient appropriate evidence to express a positive conclusion. In practice that means sampling: pick this site, pick this month, show the underlying records, show that the total reconciles, show what the controls do when data is missing.

Here is the operational consequence. A reporting process built on an annual spreadsheet can usually survive limited assurance, because the spreadsheet can be explained. It struggles badly under reasonable assurance, because the question stops being "explain your method" and becomes "show me this month at this site, from source". If the organisation cannot trace a reported tonne back to the intervals it came from, the assurance engagement turns into a forensic reconstruction billed by the hour — and the cost of the report goes up rather than down, exactly at the point the organisation hoped it was becoming routine.

What changes when the data is continuous

If consumption is already being collected at interval granularity, reconciled to sites, and quality-checked as it lands, the annual report stops being a project and becomes a query. The unit economics change in a specific way: the fixed annual assembly cost is replaced by a one-off integration cost plus a small marginal cost per reporting period. The second year is cheaper than the first, and the fifth is cheaper still, which is the opposite of the document-assembly pattern.

But the more valuable consequence is the one that happens the other fifty-one weeks of the year. The same reconciled series supports:

• Scope 2 at any period boundary, location-based or market-based, without a re-assembly exercise — and with the coverage percentage attached, so an auditor can see what it was computed from.
• Out-of-hours load analysis, which is where the recoverable cost usually is. A building drawing most of its energy when it is empty is both an emissions problem and a bill — and it is the finding that monitoring-only deployments observe and never act on.
• Verified reductions, measured against a proper baseline, so a claimed reduction survives assurance rather than resting on a year-on-year subtraction that a warm winter can manufacture on its own.
• Site-level ranking, so capital goes to the buildings with the most recoverable intensity rather than the ones with the loudest facility manager.

The double benefit, stated plainly

Regulatory reporting is becoming non-optional across most jurisdictions our customers operate in, and the direction of travel is towards assurance — meaning the number has to withstand an auditor, not merely be submitted.

So the spend is happening regardless. The question is only whether the spend also produces operational value, and that turns almost entirely on one architectural decision: whether the carbon figure is derived from continuously measured consumption, or assembled annually from documents.

Derived from measurement, the same pipeline pays for itself in avoided consumption and then produces the report as a by-product. Assembled from documents, it produces the report and nothing else, every year, forever. The asymmetry is worth sitting with: both routes produce a compliant disclosure, and only one of them produces a reason to have collected the data.

There is a second-order effect too. A reduction you can verify is a reduction you can disclose without hedging. Savings are, as the IPMVP notes, "the absence of consumption or demand" — they cannot be measured directly, only inferred by comparing before and after with adjustments for changed conditions. Organisations that never built that comparison end up claiming reductions they cannot separate from the weather, which is precisely the sort of claim an assurance provider is engaged to find.

What "assurance-ready" has to mean

A number that will survive assurance has properties beyond being correct:

• Provenance. Every figure traces to the intervals it came from, and those intervals trace to a source and an ingestion time.
• Stated coverage. If 4 per cent of intervals were missing, the report says so rather than silently totalling the remaining 96 per cent as if it were the whole year.
• Declared factors. The emissions factor, its version, its source and its effective date, attached to the calculation rather than living in a consultant's spreadsheet.
• Reproducibility. Re-running the same period produces the same number, or explains what changed.
• Exportability. The customer can extract the underlying series and check it against their own source. A platform that will not let you verify it is asking for trust it has not earned.

That last one is worth dwelling on. If the only way to check a reported figure is to ask the vendor, the figure is not evidence — it is testimony. Assurance is the process of turning testimony back into evidence, and it is much cheaper when the evidence was kept all along. This is why we treat a missing interval as an absence with a reason rather than a zero: a zero is an assertion that no energy was used, and it is almost always false.

What a measurement platform cannot do for your carbon report

This section exists because the category oversells, and an inflated claim is found out at the worst possible moment — in the middle of an assurance engagement. Candidly, then.

• Scope 3 is mostly out of reach. Fifteen categories, and metered building energy addresses essentially none of them. Purchased goods and services, capital goods, business travel, employee commuting, use of sold products: these come from procurement, HR and supplier systems. A platform that claims to produce your Scope 3 from meter data is describing an estimation model, and you should ask to see its assumptions.
• Fugitive emissions are not metered. Refrigerant leakage from chillers and split systems is a real and often material Scope 1 line, and it is established from service records, gas purchase logs and charge quantities — not from an electricity meter. Electrical consumption will not tell you a system is losing refrigerant.
• Anything without an instrument stays an estimate. A landlord-supplied heat service with no sub-meter, a tenancy apportioned by floor area, a generator with run hours and no fuel meter. These can be documented honestly as estimates, with their method recorded. What they cannot be is silently promoted to the same status as a measurement — which, in an assembled spreadsheet, is exactly what happens to them.
• We do not operate your plant. Radr measures, analyses and verifies. The decision to change a schedule, replace a chiller or renegotiate a supply contract belongs to you and your contractors. Our job is to establish what the situation is beforehand and what the change was worth afterwards.
• We do not opine on your disclosure. A platform that produces data for a report is not an assurance provider, and should not be mistaken for one.

Stating the boundary is not modesty. A carbon report whose measured and estimated components are clearly separated is a stronger document than one where everything is asserted with equal confidence, because the reader can see which parts carry weight. The same logic runs through our whole approach to measurement.

The reframe

ESG reporting is often positioned as a compliance obligation with a cost attached. It is more accurately a measurement obligation with a compliance output — and measurement obligations, once discharged properly, pay.

The organisations getting value from this are not the ones spending more on reporting. They are the ones who noticed that the data they were compelled to collect was the data they needed anyway, and who built it once instead of assembling it annually. If that is the position you are in, our ESG reporting work starts from what you already meter.