AI tax benchmarking software for Australian accounting firms

Published 27 August 2026. Last reviewed 27 August 2026. Technically reviewed by the Taxpartna tax team.

ATO benchmark checking is being added to Taxpartna. This page describes the intended review approach. The final production feature may evolve as it is tested and as ATO benchmark data changes.

ATO benchmark checking, built into tax-file review

ATO small-business benchmarks are useful because they provide an external comparison point against the information reported by similar businesses. But for an accounting firm, the value is not simply knowing that a client is inside or outside a published range.

The useful question is:

Does this result give the reviewer a reason to ask another question before sign-off?

Taxpartna is extending its completed-file QA approach to include that benchmark signal. Instead of treating the ATO range as an automatic compliance decision, the planned feature will present the variance to the reviewer with enough context to investigate the result and document the conclusion.

What are ATO small-business benchmarks?

The Australian Taxation Office publishes small-business benchmarks that allow businesses and advisers to compare business performance with other businesses in the same industry.

The ATO's software-developer material says its benchmark dataset is calculated from more than two million small-business tax returns and provides benchmark ratios for 100 small-business industries. The data is independently verified and updated annually.

Benchmarks can include ratios such as:

• cost of sales to turnover
• total expenses to turnover
• labour to turnover
• rent to turnover
• motor vehicle expenses to turnover

The relevant ratios vary by industry and turnover range. The ATO also identifies a key benchmark range for particular industries.

Why accountants look at ATO benchmarks before lodgement

Benchmark data can help a practitioner identify an unusual result that deserves more attention before the return is finalised.

For example, an unexpectedly high expenses-to-turnover ratio might be explained by:

• a genuine change in the client's business model
• unusually high input costs
• a new or disrupted business year
• large one-off expenditure
• different labour arrangements
• an incorrectly classified industry
• incomplete income reporting
• an expense-classification error
• another fact specific to the client

The benchmark does not tell the practitioner which explanation is correct. It tells the practitioner that the result may deserve a closer look.

Does being outside an ATO benchmark mean the return is wrong?

No. An ATO benchmark is not a pass/fail compliance test.

ATO guidance cautions against treating benchmark data as definitive or using it in isolation. Legitimate businesses can sit outside an industry average, and a business whose ratio is close to an industry norm can still have underlying tax issues.

That distinction is fundamental to how Taxpartna intends to use benchmarking.

A benchmark result should be presented as:

Review signal. Client ratio is outside the applicable published benchmark range. Consider whether the variance is explained by the client's circumstances and whether reported income, expenses and underlying classifications are appropriately supported.

It should not be presented as:

Not a compliance verdict. Client failed the ATO benchmark and is non-compliant.

How Taxpartna's ATO benchmark check is planned to work

Taxpartna already sits at the review stage of the tax workflow. The planned benchmark check extends that model rather than creating a separate advisory product.

1. Review the completed tax file

The relevant workpapers, financial statements, return and supporting documents are submitted through the existing Taxpartna review workflow.

2. Identify relevant business information

The review uses the information available in the submitted file to identify the client's business and the data relevant to a benchmark comparison.

3. Match the applicable published benchmark

Where an applicable ATO small-business benchmark exists, the system can use the relevant industry and turnover context to select the comparison range.

4. Calculate the ratio

The underlying benchmark ratio should be calculated using deterministic logic rather than asking a generative AI model to guess the arithmetic.

5. Compare with the published range

The client result is compared with the relevant published ATO range.

6. Surface material variances

A result outside the range is presented to the reviewer with enough context to understand what was compared.

7. Practitioner investigates and signs off

The practitioner decides whether the result is explained, whether further evidence or client enquiry is needed and whether the tax file is ready for sign-off.

Example: how a benchmark exception could appear in Taxpartna

The figures above are deliberately illustrative. Any live implementation should use current ATO data for the relevant industry and turnover band.

The business industry code matters

An ATO benchmark comparison is only useful when the business is being compared with an appropriate industry.

The ATO uses Business Industry Codes (BICs) in tax return reporting. The ATO's BIC tool notes that choosing the correct code helps place the return in the right category and can reduce the risk of a benchmark comparison based on the wrong industry.

That means industry classification itself can become part of the review question:

• Is the business industry code appropriate for the client's main business activity?
• Does the selected benchmark correspond with that industry?
• Is the turnover band correct?
• Does the client operate in a way that differs materially from the population behind the benchmark?

A benchmark calculation is only as useful as the classification and source data behind it.

AI benchmarking versus a spreadsheet benchmark check

ATO benchmark calculations are not inherently an AI problem. A ratio such as total expenses divided by turnover should be calculated exactly.

The opportunity for AI is around the surrounding review work:

• reading information spread across workpapers and financial statements
• identifying which source values appear relevant to the benchmark calculation
• connecting the result to the rest of the completed tax file
• surfacing supporting context for the reviewer
• helping the reviewer focus on exceptions rather than manually hunting for every input

Taxpartna's approach is therefore intended to combine AI-assisted document understanding with deterministic checks where exact arithmetic or comparison is required.

The calculation should be explainable. The source should be visible. The final conclusion should remain with the practitioner.

Why benchmarking belongs in the tax QA process

A benchmark check is most useful when it is not performed in isolation.

Suppose a hospitality client is outside a published expenses-to-turnover range. A reviewer may then want to look at:

• reported sales
• cost of sales
• labour
• rent
• motor vehicle expenses where relevant
• year-on-year movements
• industry code
• one-off expenditure
• trading-stock treatment
• supporting ledgers or workpapers

That investigation sits naturally beside the other questions already asked during tax-file review:

• Does the tax reconciliation agree to the accounts?
• Do the final workpapers agree to the tax return?
• Are material figures supported?
• Are expected documents present?
• Do unusual movements have an explanation?
• Are technical-risk areas ready for practitioner assessment?

Benchmarking becomes another lens over the completed file - not a substitute for reviewing the file itself.

What should a practitioner do when a client falls outside a benchmark?

The appropriate response depends on the client and the reason for the variance. A practical review process might be:

1. 1Confirm that the correct industry and turnover range have been used.
2. 2Confirm the source figures used in the calculation.
3. 3Check that the ratio itself has been calculated correctly.
4. 4Compare the current year with prior years where useful.
5. 5Look for obvious changes in the business model or trading conditions.
6. 6Consider whether income and expenses are appropriately captured and classified.
7. 7Raise a client or preparer query where the variance is not readily explained.
8. 8Retain relevant evidence or explanation in the client file.
9. 9Apply professional judgement before concluding whether any tax treatment or return amount requires action.

The benchmark should sharpen the review question, not predetermine the answer.

ATO benchmark review example

Consider a fictional small business whose final workpapers produce an expenses-to-turnover ratio outside the published range for the selected industry and turnover band.

A useful reviewer note might be:

Illustrative reviewer note. ATO benchmark exception identified. Confirmed industry code and ratio inputs. Current-year expenses are elevated due to a documented premises relocation and one-off fit-out costs. Supporting invoices and ledger detail reviewed. No adjustment concluded from benchmark variance alone; registered tax practitioner to complete final assessment.

This is more useful than simply recording "outside benchmark". It shows the signal, the investigation and the practitioner pathway to a conclusion.

Can ATO benchmarks predict an audit?

Benchmarks are among the tools the ATO can use to identify businesses whose results vary from expected industry patterns, but no software should tell a practitioner that one benchmark result predicts or prevents an audit.

The ATO itself describes benchmarks as a tool for comparison and risk identification. The practitioner should use the result as one piece of information alongside the actual facts of the client's business.

Can a business be inside the ATO benchmark and still have a tax problem?

Yes. A ratio close to an industry range does not prove that income, deductions, classifications or other tax positions are correct.

That is another reason Taxpartna should not turn benchmarking into a green-tick compliance feature. A result inside a range can be useful context, but it cannot replace the other cross-document, workpaper and tax-law checks that form part of the review.

Built for the reviewer, not for automatic tax conclusions

The planned benchmark check follows the same principle as the rest of Taxpartna:

Taxpartna surfaces the issue. The practitioner makes the decision. Benchmark data can help identify an unusual pattern. It cannot explain the client's facts, decide whether an adjustment is required or determine whether a tax return is compliant.

Where a benchmark variance appears, the reviewer may need to:

• check source records
• ask the preparer a question
• contact the client
• consider whether the industry classification is appropriate
• review a workpaper in greater detail
• obtain more evidence
• document why no further action is required
• make an adjustment where the practitioner concludes one is necessary

How this fits with Taxpartna's existing tax QA

Taxpartna is designed to review the completed tax file before professional sign-off.

It already helps practitioners examine information across tax returns, financial statements, workpapers and supporting documents and surfaces potential issues for professional assessment.

ATO benchmark checking is a natural extension of that review model because it adds an external comparison point to the evidence already in the client file.

The planned workflow is:

Completed file > Taxpartna QA > ATO benchmark signal > practitioner investigation > documented resolution > sign-off. The benchmark feature is an additional review check, not a separate tax-advice engine.

Coming soon: ATO benchmark checking in Taxpartna

Taxpartna is adding ATO small-business benchmark checking to its Australian tax QA platform.

The feature is being designed to help accounting firms:

• compare relevant client ratios with applicable published ATO benchmark ranges
• surface unusual results before tax return sign-off
• give reviewers context around the benchmark exception
• reduce manual benchmark spreadsheet work
• bring benchmark review into the same completed-file QA process as other tax checks
• retain practitioner control over every conclusion

The feature will not treat a benchmark range as an automatic compliance test. It will help the reviewer see another potential exception and decide what, if anything, needs to happen next.

See the rest of the Taxpartna review workflow

Taxpartna is built for the stage after the file has been prepared and before the registered tax practitioner signs off.

Explore how Taxpartna reviews completed workpapers, traces information across sources and presents potential exceptions for professional assessment.

Related Taxpartna resources

• Tax QA software Australia
• Tax workpaper review software
• AI tax software Australia
• Tax return review checklist
• Source hints and review traceability

Primary sources

• ATO Software Developers - Small business benchmarks
• Australian Taxation Office - Small business benchmarks
• ATO - Small business benchmarks methodology and ratio calculations
• ATO - Business industry code tool

General information only. This page does not constitute tax, legal or professional advice. Taxpartna is a quality-assurance assistance tool designed for use by registered tax practitioners. Taxpartna does not provide tax advice, tax agent services or BAS agent services, is not endorsed by the ATO or TPB, and does not make tax conclusions. All professional judgements, review conclusions and sign-off decisions remain the responsibility of the registered tax practitioner.