ato logo
Search Suggestion:

Benchmarking methodology

Our benchmarking methodology consists of 10 steps.

Last updated 13 March 2024

Step 1: Identify industries to benchmark

This step determines the industries we benchmark.

Small business benchmarks are currently limited to businesses that supply goods and services directly to consumers. Identification of an industry type to benchmark is based on a number of factors, including the size of the industry.

Industries identified for benchmarking are based on their ATO business industry codes (BIC). BICs are unique 5-digit identifiers adapted from the 4-digit Australian and New Zealand Standard Industry Classification (ANZSIC) code. To help you find the correct code, use the BIC tool. The 5-digit codes refine the industry classifications.

We allocate businesses to a benchmark industry based on both:

  • BICs
  • keywords reported at the main business activity and trading name labels on their tax returns and business registration (see step 3).

Step 2: Identify the starting population of industry business types

This step determines the starting population that are in business.

The starting population for a benchmarked industry is identified 12 months after the statutory due date for lodging the tax return.

This allows for most of the tax return and activity statement lodgments to occur, including for those businesses with extended lodgment periods. For example, the starting population for the 2021–22 financial year was identified after 31 October 2023.

Firstly, the starting population of a benchmarked industry is selected based on businesses that:

  • lodged their tax return for the year to be benchmarked
  • are registered and have an active ABN
  • report the selected BIC on their tax returns or business registrations.

Secondly, businesses are excluded from the starting population if they:

  • are currently insolvent
  • are deceased
  • are in the not-for-profit, government or large market segments
  • are not companies, partnerships, trusts or individuals
  • have a compromised tax file number (TFN).

Thirdly, further data is captured and filters applied to:

  • exclude businesses with turnovers of less than $30,000 and more than $15,000,000
  • exclude businesses that reported multiple activities on their tax returns
  • exclude those that have been in business for less than one year before the start of the financial year
  • capture the latest returns lodged for the benchmark year, including amendments
  • limit the population to those that have lodged returns within one year from the statutory due date.

Step 3: Industry allocation – group businesses into industries

This step allocates businesses to specific industry sub-groups, based on having a common business model.

We use keywords in the Main business activity and Trading name labels on tax returns and business registrations to allocate a business to a sub-group.

BICs may include more than one industry sub-group. For example, 32430 – Tiling and Carpeting Services, has 3 benchmark industry sub-groups:

  • Carpet laying services
  • Tiling services – floor and wall
  • Timber floor sanding

A benchmark industry can also include multiple BICs where similar sub-groups report within multiple codes. This is because businesses can have diverse product lines but similar financial performance. For example, the Meat and poultry retailing – fresh benchmark industry consists of:

  • Meat retailing – BIC 41211
  • Fresh poultry retailing – BIC 41213.

Industry allocation keyword process

The keyword process allocates a business to an industry sub-group based on a 3-step keyword process.

1: Tax return

Assign business to an industry sub-group population if they have reported either of the relevant:

  • BIC and keyword description at the Trading name label
  • keyword description at the Main business description label.

We also use keywords to exclude businesses that report certain descriptions at the Trading name label.

2: Business registration

Assign business to an industry sub-group population if they report the relevant information for all the following:

  • BIC in their GST business registration
  • keyword description at the Trading name label on the tax return
  • keyword description at the Main business description label on the GST business registration.

We also use keywords to exclude businesses that report certain descriptions at the Trading name label on their tax return.

3: Keyword

The keyword process identifies businesses that incorrectly reported their BICs and allocates them into an appropriate industry sub-group.

We only apply this to businesses already within a BIC selected for benchmarking. The process is not applied to all benchmark industry sub-groups.

The process can consist of up to 4 parts:

  • Allocate the businesses to an industry sub-group only when the Trading name and Main business description labels from the tax return both contain one of the key words.
  • Allocate the business to an industry sub-group only when the Trading name and Main business description labels from the GST business registration both contain one of the key words.
  • Allocate the business to an industry sub-group only when the Main business description label from the tax return and the Main business description label from the GST business registration both contain one of the key words.
  • Use keywords from the trading name or the main business descriptions to exclude a particular business from an industry sub-group.

Step 4: Calculate benchmark ratios

This step calculates the benchmark ratios using labels on the tax returns and activity statements.

For more information, see how we calculate benchmark ratios.

Step 5: Calculate the outliers

This step removes statistical outliers so the benchmarks are based on data representative of the population.

Outliers are taxpayers whose reported data is significantly different from the majority of the other observations in the population. Outliers are removed because they may be:

  • extreme cases
  • mistakes
  • not part of the industry population, for example the business misclassified itself but did not change its name.

Outliers are identified using Mahalanobis Distances, a statistical measure that examines each ratio in relation to the sample mean and the distribution of all the other ratios. Those with a high Mahalanobis Distance measure are considered to have a significant influence on the rest of the sample and are excluded as an outlier.

Step 6: Assign turnover ranges to benchmark industries

This step assigns the turnover ranges to the industry.

Benchmark industry populations are segmented into turnover groups to account for the variations in business performance that may occur due to the size, location and turnover of a business.

The turnover ranges are different for each of the benchmarked industries.

Three turnover ranges are used for most of the benchmarked industries – low, medium and high.

Benchmarked industries with smaller populations may be presented with only 2 turnover ranges.

Turnover ranges are determined by analysing the distribution of results for each benchmark industry. Industry specific factors, including clusters and trends in reporting are considered. The ranges may be influenced by:

  • variability in the distributions of results
  • the gradual increase in the average of the key benchmark ratios.

Step 7: Calculate the benchmark ranges

This step assigns the ratio ranges around the average (mean). For example, the ratio of total expenses to turnover for each range.

For every industry benchmarked, we calculate the average ratio for each turnover range. Business model, location and region affects business performance, so we determine a range around the average.

The range is represented by 30% of the population around the average. For example, a benchmarked industry with a regional area ratio of 37% and a metro area ratio of 35% would be captured in the benchmark range of 32% to 40%.

All benchmark ratios are published as whole numbers. Conventional rounding rules apply.

Step 8: Statistical assurance

This step ensures the populations are statistically valid and allows us to determine whether the ratio is a valid indicator of turnover

Normality and homogeneity testing

The statistical validity of the benchmarks is tested by checking that the benchmark populations are normally distributed and homogenous. The selection of the key benchmark ratios and the secondary benchmark ratios for publication is dependent on the outcome of the normality and homogeneous testing.

Benchmark ratios are only published if they pass these tests.

Key benchmark ratios

The key benchmark ratio is the benchmark we use to identify businesses which may not be reporting some or all their income.

We identify these ratios to make it clear to businesses and tax practitioners what benchmark we consider to be the most accurate predictor of business turnover.

For a benchmark ratio to be selected as a key benchmark ratio, the benchmark industry population must satisfy the following requirements:

  • It is normally distributed and homogenous.
  • 50% or more of the population in the industry within the turnover range is reporting the benchmarked expense.
  • There are 50 or more businesses in the industry within the turnover range.

If more than 2 tax return benchmark ratios satisfy the key benchmark ratio test, we consider the percentage of the population reporting the expense to select the most accurate predictor of turnover for an industry.

For example, cost of sales is considered an accurate determinant of sales for retailers because it is a variable cost that generally changes in direct proportion to the increase or decrease in sales.

Secondary benchmark ratios

Secondary ratios are those not identified as a key benchmark ratio. For a benchmark ratio to be selected as a secondary benchmark ratio, the benchmark industry population will satisfy the following requirements:

  • It is normally distributed and homogenous.
  • 25% or more of the population in the industry within the turnover range is reporting the benchmarked expense.

Not all benchmark ratios apply to every industry as they do not satisfy the 2 requirements given above. For example, many service entities are unlikely to have significant cost of sales.

Step 9: Quality assurance testing

A quality assurance process is conducted on both the benchmark process and the final output.

All benchmark ratios are reviewed prior to publication. This review ensures:

  • there have not been any calculation errors
  • the results are consistent with our expectation of the relevant industry.

For example, a decrease in the average of a key benchmark may indicate a change in the industry population reporting.

We may compare the Small business benchmarks to external information and other published industry benchmarks when developing a new industry benchmark.

Step 10: Industry names and overviews

Benchmarks are published by industry name and include an overview of the industry characteristics.

The main focus of the industry overview is to describe the businesses captured within the benchmarked industry.

QC81798