One Analyst, 15 Clients, 3,000 Documents: How Carbon Cons...

Your best analyst just told you she's going in-house. A property trust offered her $140,000 plus equity, and honestly, you can't blame her. She spent the last eight months typing numbers from electricity bills into spreadsheets for six different clients. That's not what she studied environmental science for. That's not what you hired her for either.

And now you're staring at your pipeline. ASRS Group 2 kicks in for financial years starting 1 July 2026. Every NGER-registered corporation - all 978 of them from the 2024-25 reporting year - gets pulled into mandatory climate disclosure automatically, regardless of whether they meet the size thresholds. Most of them have never produced a GHG inventory. They don't know the difference between location-based and market-based Scope 2 emissions. They're going to call someone like you.

But you've only got three people. And finding a fourth is going to take months - if you can find one at all.

The arithmetic that breaks consulting firms

We talk to sustainability consulting firms every week. Boutique practices, mid-tier advisory teams, independent carbon accountants. The pattern is the same everywhere. The demand surge from mandatory reporting is real, but the capacity to meet it isn't.

Here's the maths. A typical mid-market client - 20 to 40 sites, mandatory NGER reporter, now entering AASB S2 compliance - sends you somewhere between 500 and 2,000 documents per quarter. Electricity bills across multiple sites. Gas invoices. Fuel dockets from a vehicle fleet. Water statements. Waste transfer notes. Each document takes 3 to 8 minutes to process manually: open the PDF, find the consumption figure, check the billing period, enter the data, match the emission factor, file the source. Call it 5 minutes average.

At 1,000 documents per client per quarter, that's 83 hours of pure data entry. Per client. Per quarter.

An experienced analyst, working flat out, can process about 200 documents per week. So one client's quarterly data takes roughly four weeks of full-time work. If you've got eight clients, that's 32 weeks of data entry per quarter - but a quarter only has 13 weeks. You need two and a half full-time people just to type numbers. Before anyone does a single hour of advisory work.

Now imagine you've got 15 enquiries from Group 2 reporters who need help by October 2026. At 1,000 documents each, that's 15,000 documents per quarter. That's 1,250 hours of data entry. That's over five full-time equivalents doing nothing but reading PDFs and typing into spreadsheets.

Where are those five people coming from?

The talent market won't save you

They're not coming from the talent market, because the talent market is cooked.

Talent Nation's FY26 remuneration survey - covering more than 1,500 sustainability, environment, and ESG professionals across Australia - found that 54% report being stretched to the point of impact or burnout. Among those aged 25 to 30, 40% changed roles in the past 12 months. Sixty-nine percent changed within two years. Your next hire is probably already planning their exit.

And it's not just sustainability roles. Australia faces a projected shortfall of more than 10,000 qualified accountants by 2026, according to ABS demand modelling. Enrolments in Chartered Accountants ANZ's Accounting Professional Year program collapsed from 7,122 in 2018 to roughly 340 in 2024 - a 95% decline. The pipeline is broken at every level.

The World Economic Forum puts it starkly: only one in eight workers globally have skills relevant to addressing the climate crisis. Green job postings grew 22.4% between 2022 and 2023, nearly double the 12.3% growth in workers with green skills. That gap hasn't closed. It's widening.

Big Four engagement timelines for ASRS advisory work have blown out to 6 to 8 months. Some mid-market companies couldn't get a team at all for the 2025-26 financial year. Mid-tier firms like BDO, Grant Thornton, and PKF picked up overflow - and then hit their own walls.

You can't hire your way out of this. But you can change what your existing team spends their time on.

Where the hours actually go

We've written about what consultants charge in detail before. Independent boutiques bill $150 to $250 per hour. Mid-tier firms run $200 to $350. Big Four advisory sits at $250 to $500. A mid-market NGER and AASB S2 engagement - data collection through lodgement-ready output plus climate disclosure preparation - costs clients $80,000 to $200,000 per year.

But look at where those hours go. Consistently, across every consulting team we've spoken with, somewhere between 50% and 70% of billed engagement hours aren't advisory work. They're data collection.

Chasing utility bills from facility managers who never respond to emails. Opening PDFs one by one. Finding the kWh figure buried on page two. Typing it into the right cell. Cross-referencing the NGA Factors 2025 workbook to grab the correct state-based grid emission factor - 0.64 kg CO2-e/kWh for NSW, 0.78 for Victoria, 0.22 for South Australia. Filing the source document so the audit trail holds up when assurance comes around.

That work has to get done. But it's not what your clients are paying $200 an hour for. And it's definitely not what's keeping your best people engaged. 54% burnout doesn't come from interesting advisory work. It comes from the 300th fuel docket of the week.

What changes when document processing is automated

Here's the shift. Instead of one analyst opening 1,000 PDFs per client per quarter, those documents flow through an AI processing pipeline. Each document gets classified - electricity bill, gas invoice, fuel receipt, water statement. The relevant data fields get extracted: consumption, billing period, meter reference, supplier, cost. The system validates the numbers against expected ranges for that site. Emission factors from the NGA Factors database get matched automatically, including state-specific grid factors and the correct scope allocations.

The consultant's job moves from data entry to data review. You're not typing "4,327 GJ." You're checking that the system got it right, flagging anomalies, making judgement calls where the data is ambiguous. A bill that's 3x the usual consumption for a site. A fuel type that doesn't match the fleet records. A missing billing period.

That review process takes a fraction of the original time. What was 83 hours of data entry per client per quarter becomes 8 to 12 hours of review. Across 15 clients, that's 120 to 180 hours of review - well within reach for a three-person team - instead of 1,250 hours of data entry that would need five full-time analysts.

We should be honest: the system doesn't get everything right. Handwritten fuel dockets with faded thermal ink remain difficult for any processing system, human or AI. Merged PDF statements where three billing periods are combined on one page cause ambiguity. Unusual utility formats from niche suppliers sometimes need manual handling. We're still working through edge cases with some document types, and we won't pretend we've solved them all.

But for the 85 to 90% of documents that are structured utility bills, standard fuel receipts, and machine-printed invoices? The extraction accuracy is high enough that review-by-exception is a genuine workflow, not a marketing claim.

The financial model for a consulting firm

Let's make this concrete. Consider a boutique consulting firm. Two senior consultants, one analyst. Currently serving six clients.

Before - the data entry bottleneck:

Each client generates roughly 800 documents per quarter. At 5 minutes per document, that's 67 hours per client per quarter, or 400 hours across six clients. The analyst handles most of it, but even working full-time on data entry (480 hours per quarter), she can barely keep up with six clients. The seniors fill the gaps, pulling them away from advisory work. Total annual revenue at $65,000 per client: $390,000.

The seniors want to take on more clients. They can't. The bottleneck isn't their advisory capacity - it's the data processing that precedes every piece of advisory work. And they won't hire another analyst at $105,000 to $142,500 fully loaded (the real cost, once you add 12% super, leave loading, payroll tax, and workers' comp) because the firm can't guarantee enough new clients to justify the headcount.

After - advisory-focused delivery:

Same three people. AI-powered document processing handles the extraction and factor matching. The analyst now spends 10 hours per client per quarter on review and exception handling instead of 67 hours on data entry. That's 60 hours across six clients - leaving the analyst 420 hours per quarter for higher-value work like Scope 3 data collection, draft report preparation, and assurance documentation.

More importantly, the firm can now take on 12 to 15 clients. At a lower per-client fee - say $45,000 instead of $65,000, because the engagement is more efficient - twelve clients generate $540,000. Fifteen generate $675,000. Same headcount. Revenue per person nearly doubles.

And the engagement quality goes up. Every advisory hour that used to be swallowed by data entry is now available for the work clients actually need: interpreting results, building reduction strategies, running AASB S2 scenario analysis, preparing transition plans, training boards on climate risk.

The software subscription - somewhere between $15,000 and $60,000 annually depending on volume - is a fraction of one FTE salary. The maths isn't close.

What the workflow actually looks like

We built Carbonly with consulting firms as a core use case. Our founding team spent 18 years inside organisations like BHP, Rio Tinto, and Senex Energy, working alongside consultants on exactly this kind of engagement. We know the pattern. We built the platform to fit it.

Multi-client isolation. Each client exists as a separate organisation within the platform. Separate projects, separate access controls, separate audit trails. A consultant managing twelve clients sees a portfolio dashboard - "Client A: 147 records processed this quarter, 6 need review. Client B: 203 records, 2 flagged." - without any data leaking between clients. This isn't a nice-to-have. For consulting firms, it's table stakes.

Continuous document ingestion. Each client project gets a dedicated email address. Facility managers forward their utility bills as they arrive throughout the quarter. No more "please send us all your bills by the 15th" emails in September. No more three-week waiting periods. Documents flow in, get processed automatically, and sit ready for the consultant's next review.

Exception-based review. The consultant doesn't check every record. They check the ones that need human judgement: documents with low confidence scores, consumption figures outside expected ranges, new emission sources that weren't in last quarter's data, unrecognised suppliers. Professional judgement where it matters - not rubber-stamping every kWh figure.

Pre-built reporting. When lodgement time arrives, NGER category summaries and AASB S2 disclosure inputs are already structured from clean, validated data. The consultant isn't rebuilding a spreadsheet from scratch. They're reviewing and refining a pre-populated output.

The cross-client advantage nobody talks about

When you manage multiple clients on the same platform, something interesting happens. You start seeing patterns across your portfolio that are invisible when every client lives in a separate spreadsheet.

Which of your NSW clients has Scope 2 intensity per square metre more than 20% above the cohort average? Where did diesel consumption spike quarter-on-quarter, and does it correlate with a site expansion or a data entry error? Which clients are still using the national average grid factor (0.62 kg CO2-e/kWh) when they should be using their state-specific one?

That's not just efficiency. It's a new kind of advisory insight. You can benchmark clients against each other - anonymised, obviously. You can spot a billing anomaly at one client because you know what normal looks like across ten others. You can identify a reduction approach that worked for one client in construction and recommend it to another.

We're not sure this scales cleanly for consultants managing 50-plus clients across wildly different industries yet. The benchmarking becomes less meaningful when you're comparing a property trust to a mining operation. But for a boutique firm running ten to twenty clients in similar sectors? It changes the advisory relationship from reactive reporting to proactive insight.

This isn't about replacing you

We want to be direct. AI tools don't replace sustainability consultants. Not for the work that actually matters.

Deciding whether a subsidiary falls under operational control or financial control for NGER purposes - that's regulatory judgement. Determining the right scenario analysis methodology for a mining company versus a property trust under AASB S2 - that's strategic thinking. Advising a board on the financial implications of the Safeguard Mechanism's 4.9% annual baseline decline - that's expertise. Interpreting whether a client's net zero claim will survive ACCC scrutiny - that's professional risk assessment.

None of that is automatable. And none of it should be.

What is automatable is the 50 to 70% of engagement hours currently consumed by reading utility bills, matching emission factors, calculating Scope 2 emissions, maintaining audit trails, and generating draft NGER summaries. That work is necessary. But it doesn't require your expertise. It requires a system.

The consultants who adopt these tools aren't making themselves redundant. They're reclaiming capacity for the work their clients are actually paying for. They're moving up from data processor to strategic adviser - where they should have been all along.

The referral model that builds stickier relationships

There's a commercial angle here worth considering. If you set up a client on a carbon accounting platform for their ongoing data management, you've created a relationship that outlasts any single engagement.

The client collects and processes their own utility data throughout the year. The platform maintains the emission factor library, applies calculations, keeps the audit trail. But the client still needs the consultant for interpretation. For strategy. For regulatory advice. For assurance preparation. For the board presentation where the CFO asks "what does this mean for our capex plan?"

They just don't need the consultant for data entry anymore. And neither party misses that.

The consultant becomes the expert layer on top of the data system. The client pays less overall (because they're not funding 500 hours of data processing) but more per advisory hour (because every hour is high-value). The consultant earns more per person because every hour billed is advisory work, and they're serving three or four times as many clients.

That's a healthier business model. For everyone.

Where to start if you're at capacity already

Don't try to automate your entire client portfolio at once. Pick one client - ideally one with a messy document stack and a predictable quarterly cycle. Upload their last quarter's documents in bulk. Compare the AI output against what your analyst would have produced manually. Check the emission factors. Check the state-based grid factors. See where it nails the structured electricity and gas bills (it usually does) and where it needs help (handwritten dockets, merged statements, unusual formats).

That first client gives you confidence. By the third, your team will have a rhythm. By the eighth, you won't remember how you ever ran the old way.

The Group 2 wave starts in three months. If you're a consulting firm already at capacity with Group 1 clients and your existing NGER book, the window to get your workflow sorted before the phones start ringing is closing fast. Every week you spend manually processing documents for current clients is a week you can't spend onboarding new ones.

The talent market isn't going to fix this for you. But your existing team, with the right tools underneath them, might be enough.

Related reading:

Carbonly.ai is built for consulting firms as much as for in-house teams. Multi-client management with full data isolation, AI document processing that handles the extraction your analysts currently do by hand, NGER-native reporting, and AASB S2 disclosure outputs - all from a single platform. If you're a consulting firm staring at a pipeline you can't service with your current headcount, we'll process a client's quarterly documents for free so you can see whether the workflow fits.