The cost conversation about climate disclosure usually starts in the wrong place. It starts with the size of this year’s invoice, when the question that matters is the slope of the line. Spend on Australian Sustainability Reporting Standards (ASRS) compliance is rising for every captured entity, and it will continue to rise as the assurance bar steps up. But rising spend and rising assurance are not the same line, and that is the trap. A resource company can spend more every year and still reach the reasonable assurance cycle with nothing for an assurer to test. The current model’s cost trajectory does not scale to reasonable assurance because it buys a deliverable rather than a capability.
This is the piece for the CFO who wants the cost question modelled rather than guessed. It sets out the anatomy of resource-sector climate disclosure spend between FY24 and FY26, where consultant cost reduction actually lands once you look past the headline, a total cost of ownership comparison between in-house infrastructure and continued consultant dependency, what headcount does under reasonable assurance, and the cost line almost every status-quo budget leaves out: the price of a qualified opinion. The numbers below lean on the regulator’s own published estimates, not on a vendor’s. The conclusion they point to is uncomfortable for the prevailing model, and it is the whole point of the benchmark.
Start with the regulator’s own figures, because they set the frame. Treasury’s Policy Impact Analysis on mandatory climate-related financial disclosures put the initial transition cost of the regime at between one million and 1.3 million dollars per year per entity, falling to an ongoing cost of between 500,000 and 700,000 dollars per year per firm once practices stabilise. Those are economy-wide averages across roughly 1,800 captured entities, so they understate the resource-sector position rather than overstate it. A coal or iron ore producer whose Scope 3 total is dominated by Category 11, the use of sold products, under the GHG Protocol fifteen-category structure already used for NGER reporting, sits at the top of that range, not the middle of it.
Now look at the shape of the spend, not just the size. In FY24, climate disclosure was voluntary for most resource companies. Treasury notes that disclosing under the lighter Taskforce on Climate-related Financial Disclosures framework was already costing firms between 250,000 and 400,000 dollars a year and providing them with less assurance that they were reporting material risks correctly. That spend was overwhelmingly consultant-led: a provider engaged in the months leading up to the report to assemble an estimate and produce a defensible number. By FY26, with the first AASB S2 cycle live, the figure is materially higher, but the composition has barely moved. The money still buys consultant hours, internal preparation time and an assurance fee, in roughly that order. The line item grew. The thing it purchased did not change. That is the FY24 to FY26 story in one sentence: a bigger cheque for the same deliverable, with the assurance bar still ahead of it rather than behind it.
It is worth noting why the bar is still ahead. In the first mandatory cycle, the FY26 spend is buying only limited assurance, and only over a narrow set of disclosures. Under ASSA 5010, year one requires limited assurance over Scope 1, Scope 2, governance and selected strategy paragraphs, with Scope 3 still outside the assured perimeter. So the FY26 cheque, large as it is, is being written against the easiest version of the obligation the company will ever face. The mistake the cost line invites is to read a manageable FY26 number as evidence that the model works, when in fact it has not yet been tested against anything demanding. The price rises from here while the scope it addresses widens, and a budget built by extrapolating the FY26 figure at a steady rate will understate both.
The reflex response to a rising cost line is to cut consultant spend, and the reflex is half right. The trap is assuming that the savings come from removing expertise. It does not. Methodology judgement, materiality determination, the conversation with the assurance provider and the directors’ sign-off all still need senior people, and trying to economise on those is how a company manufactures a restatement. The expertise is not the waste. The re-derivation is.
Under the year-end model, a large share of the fee pays a provider to rebuild the dataset from scratch every cycle, re-collecting the fuel invoices, the rail consignment notes, the bills of lading and the meter reads, re-keying them, and reconstructing a number that existed last year and will have to be reconstructed again next year. That is the cost that compounds and never retires. When primary data is captured at source and held in a controlled, version-controlled system, the rebuild disappears, and the expert is freed to spend their hours on judgment rather than on data archaeology. The savings lie in eliminating repeated reconstruction, not in eliminating advice. It also produces a second-order benefit that the budget rarely captures. Under ASIC Regulatory Guide 280, directors need the key judgements behind material disclosures to be documented and supportable from the company’s own records. A controlled trial delivers that as a by-product. A consultant’s working papers do not. This is the infrastructure argument, which our piece on building audit-ready emissions data infrastructure works through in full. The point here is narrower: the consultant line falls furthest when you stop paying anyone to rebuild the same dataset twice.
A fair comparison spans four reporting cycles, because that is the window in which the assurance bar moves from limited to reasonable under ASSA 5010, and a one-year snapshot flattens the model that is cheaper today and ruinous later. Set the two paths side by side.
The continued-dependency path looks cheaper in year one because there is no build, only a fee. But the fee does not hold flat. Each cycle, the scope widens as more Scope 3 categories are brought into assured disclosure, and the assurance bar itself rises. Reasonable assurance runs roughly 30 to 50 per cent more than limited assurance for a comparable scope, reflecting the greater testing, the more senior practitioners and the site work a positive opinion demands. The step up in the bar and the widening of scope arrive together, so on the dependency path, the line climbs in both dimensions at once, and the re-derivation it pays for gets more expensive as the bar rises rather than cheaper.
The infrastructure path inverts the curve. It costs more in year one, when the system is stood up, and less in every cycle after, because the run-rate replaces the rebuild. More importantly, it bends the assurance fee curve down. The encouraging half of the industry estimate is that the delta between limited and reasonable narrows as data systems mature, because mature systems provide the assurer with evidence to test rather than gaps to investigate. The dependency path can never capture that, because it presents the assurer with a freshly rebuilt number and no operating history. The crossover, where cumulative infrastructure cost falls below cumulative dependency cost, typically occurs before reasonable assurance bites, which, under ASSA 5010, applies to all mandatory climate disclosures for periods from 1 July 2030. The candid version is that the two paths cost roughly the same to reach a working, reasonable-assurance process. The difference is that one of them also pays for a remediation, and the other does not. Our cost benchmark report models the crossover for a representative Group 1 producer; the assurance trajectory underlying it is the subject of our piece on the limited-to-reasonable assurance roadmap.
Reasonable assurance tests whether controls operated across the whole reporting period, not whether the year-end number looks right. That single requirement reshapes the headcount question. A control that an assurer can rely on has to run on a monthly cadence and leave evidence that it ran, which means the work moves out of a once-a-year sprint and into the ordinary rhythm of the close. Under the consultant-dependency model, the implied staffing answer is to add internal people to feed and check the annual rebuild, on top of the external hours, and to keep adding them as the scope widens. The full-time-equivalent line rises with each cycle because more disclosure under the same manual process requires more hands.
Infrastructure does not make the headcount line vanish, and it is worth being honest about that rather than promising a function that runs itself. What it does is change the job. The role shifts from data wranglers who assemble and reconcile spreadsheets to controllers who own a process and review its output, the same shift finance made when the month-end close replaced the year-end scramble. The FTE curve flattens because the system absorbs the volume that headcount would otherwise have to absorb, so adding a Scope 3 category becomes a configuration change rather than a hiring round. The CFO who models headcount honestly will find that continued dependency does not avoid the people cost; it defers it and then compounds it, while the controlled path pays for capability once and staffs to supervise it rather than to perform it.
Every status quo budget shares the same blind spot. It costs the invoice and ignores the tail. The tail is a qualified or modified opinion on the climate disclosures, and it is exactly the outcome of the year-end rebuild courts, because a rebuilt process has no operating effectiveness for a reasonable assurance engagement to test. When that opinion lands, the costs that were never in the model arrive at once. There is the remediation itself. There is the restatement exposure if the prior number cannot be stood behind. These are the questions a qualified opinion invites at the next directors’ and officers’ insurance renewal, given that under the Corporations Act and RG 280, the board carries a reasonable-basis obligation for what it signed. And there is the capital-markets cost, which is the largest and the least visible: a sustainability-linked loan margin step-up or a wider bond spread when the data behind a covenant cannot be assured to the lender’s standard, the subject of our companion pieces on sustainability-linked loan and green bond data assurance.
None of these sits on a consultant’s invoice, which is precisely why the status quo case looks cheaper than it is. They are real costs of a model that cannot clear the bar it is heading toward, and they are correlated, because the same control weakness that produces the qualified opinion is the one that unsettles the insurer and the lender. A useful way to price the status quo is to take its visible annual spend and add the probability-weighted cost of the opinion it is exposed to. Do that, and the gap between the two paths stops being a question of this year’s fee and becomes a question of risk the board is carrying without having decided to do so.
Put the pieces together, and the benchmark is not really about who has the lowest invoice this year. The regulator’s own numbers say the regime costs most captured entities between 500,000 and 700,000 dollars a year once it settles, and a resource company carrying a material Scope 3 total sits above that. That spending is happening either way. The only live decision is what it buys. The dependency path buys a deliverable that must be rebuilt every cycle, becomes more expensive as the assurance bar rises, defers and then compounds the headcount cost, and leaves the qualified-opinion tail uncovered. The infrastructure path buys a capability that runs on the same controls discipline finance already owns, flattens the assurance fee and headcount curves, and retires the tail. Both cost real money. Only one of them compounds in your favour. This is the cost face of the same shift our synthesis piece on the CFO climate disclosure framework treats as a whole: climate reporting has moved inside the financial reporting perimeter, and the spend question follows it there.
If a benchmark of your current spend against this model would be useful, we are happy to build it with you and your finance team, using your own cost lines rather than a generic average. The full ASRS Group 1 cost benchmark report, with the underlying data, is available on request. There is no urgency engineered into the offer. The urgency, to the extent there is any, sits in the phasing calendar, not in this conversation.
The invoice tells you what this year costs. The benchmark tells you what next year is already committing you to.
Nick Ogle has over 30 years of experience in Enterprise IT, spanning engineering, sales, and marketing roles across Australia, the USA, and APJ for various IT vendors.
Nick is passionate about Entrepreneurship and Software innovation that drives positive change. Currently, he is the Sales & Marketing Manager at SCIAR Systems, a Newcastle-based SaaS startup, where he is helping to commercialise its groundbreaking Bulk Commodity Logistics & Emissions Certification solutions.