A Day in the Life of an Australian Sustainability Manager in Year One of AASB S2
ASRS Group 2 went live on 1 July 2026. The Sustainability Manager is the person actually doing the work. Here is what an hour-by-hour day in Q3 of year one looks like when 40+ subcontractor invoices arrive as PDFs, phone photos and inconsistent CSVs, and the CFO wants to know how you are tracking against last year.
The ASIC surveillance letter is not addressed to the CFO. Not the Chair. Not the assurance partner. It is addressed to the Sustainability Manager, because they signed off the Basis of Preparation, chose the emission factors, approved the supplier data, and told the CFO the numbers were ready for the audit committee.
ASRS Group 2 went live on 1 July 2026. Every NGER reporter got pulled in automatically through the registration pathway. Every entity clearing two of three thresholds ($200M+ revenue, $500M+ gross assets, 250+ employees) is now in. Q1 of the first reporting year is closed. Q2 is being finalised. Q3 is the middle of the fight, which is why we are picking Q3 for this walk-through.
What follows is a realistic Q3 Wednesday for the Sustainability Manager at a hypothetical mid-market Australian construction and infrastructure business. Around 900 people, twelve active projects, NGER Group A facility reporter, subject to AASB S2 in year one, ASSA 5010 limited assurance on Scope 1 and 2. Around 40 recurring suppliers issue emission-relevant documents every month: Ampol, BP Australia, AGL, Origin, Boral, Holcim, InfraBuild, Cleanaway, Veolia, Toll and a long tail.
7:30 AM. Overnight extraction queue check
The Sustainability Manager opens the extraction queue from their phone before leaving the house. Overnight, the AI Document Engine has processed 63 documents that landed in the per-project ingestion inbox between 6 PM and 6 AM. Every project has its own dedicated email address, plus a folder synced from OneDrive that site engineers use when the internet is patchy on a highway job.
Of the 63 documents, 58 came through as Green in the Match Provenance badge. That is the "template + high confidence + no anomaly" state. Six sit at Amber (auto-derived template applied but a field is below the confidence threshold), and one is Red (no template, low confidence, needs a human review). The Sustainability Manager filters to Red and Amber. That is a seven-item review list. In year zero, before extraction templates existed, the review list on a Wednesday morning was routinely 200+.
The Data Health Agent has also posted three overnight flags: a Scope 2 electricity account with a billing period that overlaps the previous month's invoice, a diesel volume that jumped 34% week-on-week for one project, and a supplier PDF where the ABN did not match any known vendor. Each flag links back to the source document with the exact page and coordinate highlighted. The Sustainability Manager triages: overlap is a re-issued AGL invoice for a corrected meter read, the diesel jump is real (a new civil crew mobilised on-site), and the ABN mismatch is a new subcontractor no one told them about. Ten minutes of work. In the old spreadsheet workflow, that ABN mismatch would have been discovered in October by the auditor.
8:15 AM. Site manager forwards yesterday's Ampol fuel dockets from a truck cab
At 8:14 AM the phone buzzes. A site manager on a Newcastle highway upgrade has forwarded 22 photos of Ampol fuel dockets from yesterday's shift. Some are creased. Two have coffee on them. One is a photo of a photo. All are addressed to the project's dedicated ingestion inbox, so they land straight in the right project workspace with the correct cost centre and the correct facility ID for NGER.
The Ampol dockets go through the per-supplier extraction template the platform learned after the first five clean samples: litres, fuel grade, docket number, date, site code. Twenty of the 22 clear the confidence threshold. Two are legible enough to extract but the docket number is partially obscured. The manager confirms the number from the sequence of the other 20 (the Data Health Agent already flagged the likely value). Total time: under three minutes for a stack that would have taken an afternoon to key in. This is the industry data-entry pattern the 10,000 fuel receipts a quarter benchmark is built around.
9:00 AM. The AGL invoice that broke the extraction template
AGL changed the layout of their commercial electricity invoice on 1 August. The first three invoices under the new layout arrived last week. On the fourth, the platform's per-supplier extraction template auto-derives the new field positions and re-locks. The mechanism requires five clean samples with at least 80% column recurrence before the new template supersedes the old one. This morning the fifth invoice has arrived.
The Sustainability Manager reviews the auto-derived template in the review pane. Four fields on the new layout have moved: the meter number is now in a header block instead of a footer, the daily average kWh is a new field, the emission factor reference disappeared (AGL used to print the factor themselves, now they do not), and the billing period date format switched from DD/MM/YYYY to DD Mon YYYY. The platform has correctly captured all four changes. The manager clicks Approve. Future AGL invoices from all 12 projects will apply the new template automatically, with the self-learning loop improving each subsequent match.
In year zero this would have been a spreadsheet rebuild. Someone rewriting a VLOOKUP against the new column position. Someone else discovering the change three weeks later when their pivot table returned zeros.
10:30 AM. CFO asks "how are we tracking against last year"
Slack notification. The CFO wants a comparison of Q1 FY26 vs Q1 FY25 emissions by scope, by project, in the next 40 minutes, for a private conversation with the Chair. In the old workflow this was a two-day exercise involving a re-baseline from a spreadsheet extract, a manual grid factor lookup for the previous financial year, and a lot of prayer that the pivot tables would agree with the source data.
The Sustainability Manager opens the Variance Explanation Agent and points it at Q1 FY26 vs Q1 FY25. Sixty seconds later they have a decomposition: total emissions up 9.2%. Of that, 6.1 percentage points is activity growth (four new projects mobilised). Two points is a factor version change (NGA 2025 factors are used for FY26; NGA 2024 was used for FY25, and Victoria's grid factor rose from 0.77 to 0.78 kg CO2-e/kWh). One point is a category mix shift toward higher-intensity diesel plant. The remaining 0.1 points is boundary consolidation from a small acquisition.
Factor version pinning matters here. Because every FY25 record is pinned to NGA 2024 factors, the year-on-year comparison is like-for-like at the activity layer and separately quantified at the factor layer. The Chair gets a chart, three bullet points and a source-document link. The Sustainability Manager gets 38 minutes back.
11:30 AM. Boral concrete docket layout has changed
A Data Health Agent alert: the extraction confidence on the last three Boral concrete dockets from the North Sydney project has slipped from 96% to 71%. The manager clicks in. Boral has moved the delivery ticket number from the top-right corner to a new footer position and has added a QR code that partially overlaps the volume field. Two of the three dockets have partial volume reads.
The manager runs the "propose new template" flow. The AI Document Engine has already staged the candidate template. It needs two more clean samples to lock. The manager routes the two ambiguous dockets into the Copilot review pane, confirms the m³ values against the site diary, and flags the Boral supplier record for the next batch. The supplier portal will automatically nudge Boral's account team if the new layout keeps producing sub-threshold confidence for another week. Not because Boral needs Carbonly, but because a subcontractor issuing 400 dockets a month is worth a conversation.
12:30 PM. Working lunch: the Cleanaway waste invoice does not add up
The waste invoice from Cleanaway for the Melbourne project shows a tonnage figure of 47.6 tonnes and a bin-lift count of 84. The custom calculation rule on the waste stream cross-checks the two numbers against a supplier-specific bin size and rejects anything more than 15% out of range. This month's invoice is 22% out of range. The rule surfaces the record for manual review with a plain-language reason string attached.
The Sustainability Manager pulls up the invoice source doc from the audit trail. There is a line item they missed on the first pass: a one-off skip bin hire that inflated the tonnage without a corresponding lift count. The manager splits the record using a custom formula (skip tonnage carried on a different waste category), re-runs the calculation, and both figures now sit inside the tolerance. The typed audit event captures who split the record, when, why, and against which formula version, all retained for seven years.
1:30 PM. Auditor sample request: five records to walk-through
The ASSA 5010 assurance provider is midway through their limited assurance procedures. They email five records they want to walk-through end-to-end for Scope 1 fuel and Scope 2 electricity: source document, extraction, factor version, calculation rule, approver, and any restatement history.
The manager opens the Auditor Workspace, filters to the five sample record IDs, and clicks Export Evidence Pack. Two minutes later there is a zip file containing the original PDFs, the extraction field trace, the factor version pin (NGA 2025 for FY26, edition and effective date printed on every calculation), the exact calculation formula used, the approver's typed audit event, and the source-document hash chain. The manager attaches it to the auditor's ticket and moves on. This is one of three questions an assurance provider asks that a spreadsheet cannot answer.
In year zero, "walk me through five records" was a five-person full-day exercise. Sometimes it was three days.
3:00 PM. Scope 3 supplier engagement: 12 of 20 have responded
Twelve of the 20 top-spend suppliers have responded to the Scope 3 data request that went out via the Supplier Portal three weeks ago. The Data Chase Agent has been sending scheduled reminders. Of the 12 responses, nine came in through the portal directly (structured, immediately usable), two came in as PDFs uploaded to the portal, and one came in as an email attachment routed via the per-project ingestion address.
The manager reviews the Trust Graduation state of each supplier. Two suppliers who have been reporting cleanly for three quarters have graduated from "verify each submission" to "spot check monthly". That reduces the review workload by an hour a week without weakening the audit position, because the graduation event itself is captured as a typed audit event with the rationale attached.
The eight suppliers who have not responded get a personal follow-up. Not a system-generated one. Category 1 purchased goods is where limited assurance is heading in year two under ASSA 5010, so this is the right hour to invest.
4:00 PM. Board pack question: why did Q2 refrigerant emissions spike
The board pack goes out on Friday. The Chair has asked for a note on why Q2 refrigerant emissions were 41% higher than Q1. The Variance Explanation Agent decomposes it in 90 seconds: two large split-system replacements at the head office fit-out released roughly 4.2 kg of R-410A (GWP 2088 under AR5, GWP 2256 under AR6). Under AR5 that is around 8.8 tonnes CO2-e. Under AR6 it is around 9.5 tonnes.
Because NGER uses AR5 for the FY26 report but AASB S2 disclosure applies AR6, both numbers need to appear in the board pack. The Copilot generates the paragraph, links to the invoice from the HVAC contractor as the source, and cites the exact GWP values used with their factor version pin. The Sustainability Manager reviews and drops it into the board pack draft. Twelve minutes.
5:15 PM. Evening handover to the on-call reviewer
A part-time reviewer takes overnight and weekend document review coverage. The Sustainability Manager writes a two-line handover note in the platform. The on-call reviewer's dashboard filters automatically to the 11 records currently sitting in Amber or Red across all 12 projects. The MCP-connected Copilot can be asked in natural language what changed today and returns the answer with links to every affected record. The manager closes the laptop at 5:30 PM.
What the manager did NOT have to do
The list of things that never entered the Sustainability Manager's day is longer than the list of things that did. They did not look up NGA Factor editions (factor version pinning did that). They did not convert litres to kilolitres or kWh to MJ (unit normalisation did that). They did not maintain a per-supplier column mapping spreadsheet (extraction templates did that, auto-derived after five clean samples). They did not chase source documents through email folders (every record links back with page and coordinate reference, retained for seven years). They did not rebuild a VLOOKUP after AGL's layout change. They did not manually reconcile bin-lifts against tonnage.
None of this is glamorous. All of it is the difference between a Sustainability Manager who can look the auditor in the eye and one who cannot.
How this day looked before the platform: the 12-hour version
Before the AI Document Engine, this Wednesday plays out very differently. 7:30 AM is triage of a shared inbox with 63 unread emails. 8:15 AM is hand-keying 22 Ampol dockets into Excel. 9:00 AM is a call from finance because their pivot table zeroed out on AGL after the layout change.
The CFO comparison slips to Friday and uses FY25 factors for both years because rebasing is too painful. The Cleanaway anomaly is discovered by the auditor in October. The audit sample request takes three days. The refrigerant question quotes the wrong GWP. The day ends at 8:30 PM. Compressing that 12-hour manual day into a 9-hour supervised day is the actual return on this category of software.
FAQ
What does a Sustainability Manager do in year one of AASB S2?
In year one of AASB S2, the Sustainability Manager owns operational execution of the ASRS Group 2 program: data collection from every project and supplier, factor version pinning, Scope 1 and 2 calculation under limited assurance, Scope 3 engagement across top-spend suppliers, the Basis of Preparation document, response to ASSA 5010 auditor sample requests, variance explanation for the board, and reconciliation between the NGER submission (AR5 GWPs) and the AASB S2 disclosure (AR6 GWPs).
How much time does a Sustainability Manager spend on data entry?
A manual spreadsheet workflow at a mid-market construction reporter can consume 60% to 80% of a Sustainability Manager's week during peak reporting months. The 10,000 fuel receipts benchmark is a real industry data-entry pattern for a construction business over a single quarter. With an AI Document Engine handling per-supplier extraction, unit normalisation and factor lookup, that workload compresses to an hour or two a day of review.
Who is responsible for AASB S2 sign-off in an ASRS Group 2 entity?
Directors are legally responsible for the sustainability report under the Corporations Act, and s 180 duties of care and diligence apply. The CFO typically signs the Basis of Preparation. The Sustainability Manager owns the underlying data, factor choices and methodology. The ASSA 5010 assurance provider expresses an opinion (limited assurance in year one for Scope 1 and 2). ASIC surveils the disclosure once it lands.
What software does a Sustainability Manager actually use during the day?
On a well-instrumented day the Sustainability Manager works from one carbon accounting platform: emissions ledger, extraction, factor pinning, calculation rules, Auditor Workspace, Supplier Portal and Copilot. It is fed by per-project email ingestion, OneDrive or SharePoint folder sync, and a Copilot that lets ChatGPT or Claude query the live ledger in natural language. The Sustainability Manager should not be running a spreadsheet as their primary system of record in year one of AASB S2. If they are, the assurance provider will find it.
How does the Sustainability Manager work with the assurance provider?
The Sustainability Manager is the primary counterparty during ASSA 5010 fieldwork. Practical hygiene: agree the sample selection method up front, publish the Basis of Preparation before fieldwork, link every source document to its ledger record with a hash chain, pin every calculation rule and factor version, and use an Auditor Workspace with an Evidence Pack export so "walk me through five records" is a two-minute answer, not a three-day scramble.
Related reading
- Per-supplier extraction templates for Boral, Ampol and AGL
- AI carbon accounting with self-learning and human-in-the-loop accuracy
- Variance explanation for board emissions analysis
- Three questions your assurance provider will ask that a spreadsheet cannot answer
- Supplier Portal for Scope 3 data collection
- Seven-year audit trail with version control
What this costs
Carbonly is priced per project with a $100/month workspace minimum. A mid-market ASRS Group 2 reporter with a dozen active projects lands well under the cost of one week of Big 4 assurance prep. Want the walk-through against a real Boral docket, AGL invoice and Ampol fuel receipt from your own portfolio? Email hello@carbonly.ai.