# How This Statistics Resource Was Compiled

Australian Web Experts compiled published evidence to help Australian owners and writers identify what a digital statistic measures. This is a curated resource, not an AWE survey or a national website audit. Sources were checked on 1 October 2026; the separate reference periods in the ledger describe when the observations took place.

## Selection and Provenance

We selected evidence relevant to Australian business counts, reported website/app/social-media/AI use, and search behaviour. We preferred original statistics, methods and platform documentation. We used ASBFEO's business count as official curation and reconciled it against its underlying ABS table. Each record identifies that distinction.

The downloadable ledger has seven usable source records. It retains the original claim, source URL, publisher/release or update date, observation period, geography, population/denominator, method, limits, exact source location and AWE check/review date. It contains both statistical claims and clearly labelled survey-method/technical guidance records. It is not seven comparable national percentages.

Selected Australian shopper, US browsing and platform studies have narrower applications. They appear in the resource with qualifications immediately beside their claims and are excluded from the usable-only download. We have omitted an unverified latest-period AI-tracker rate. A source being accessible does not make every possible interpretation valid.

## Checking the Numbers

ABS counts and digital estimates were checked against the original XLSX cells. Workbook versions and SHA256 hashes are retained in AWE's review evidence. The small-business count is a sum of published ABS categories, not newly collected data. Survey rates were not multiplied by administrative business counts, averaged across incompatible populations or converted to website-only absence rates.

The ABS business survey covers employing entities within its defined scope. Its size categories and exclusions remain with the figures. Estimates have sampling and reporting error. The cell-specific uncertainty for the website/app and AI values was not established from the public materials checked; whole-survey sample/response details are not cell-level denominators.

For platform summaries, we recorded the publicly disclosed sample, period, metric and sponsor, and named missing method details. Observational comparisons do not establish a causal effect on an individual business. Overseas data remains labelled overseas; subscriptions, clicks and enquiries remain different outcomes.

## Recommendations and Updates

The diagram and owner worksheet are AWE's explanations and practical tools. They are not measured outcomes. The worksheet is unfilled except for suggested decision categories; no business assessment or result is implied. Do not enter customer records, credentials or private information into it.

We check the evidence again before publication, when a relevant publisher revises its release, and on the listed review dates. Those dates are workflow reminders; they do not change the original observation periods. Keep a change record with claim ID, old/new source period, correction reason and check date. Publishing and update dates should reflect actual release work.

## Suggested Attribution

Name the original publisher and study, its observation period and population. Where helpful, identify this page as a compilation by Australian Web Experts and link to the original source. Using a figure does not require a particular link or imply affiliation with its publisher.