How Accounting Firms Get Named in AI Answers

What Does Google's Top-Firms Answer Draw On?

Google's AI Overview for "top accounting firms australia" draws on rankings coverage, not firm websites. On 19 August 2026, the answer used LinkedIn posts about the Australian Financial Review (AFR) Top 100 and firms' own ranking posts. Other sources were trade and list sites, such as consultancy.com.au and accountantsdaily.com.au. No firm service pages appeared.

This finding comes from Intelligent Resourcing's own check of that search. It is one snapshot of one answer, and AI Overviews change each time they run.

The pattern still matters, because it shows which kinds of page the engine trusts for this question. That makes it the starting point for content built for AI citation.

The journalist row reflects a wider pattern. Kai-Cheng Yang's 2025 study of more than 366,000 AI search citations found 9% pointed to news outlets.

AFR journalist Edmund Tadros covers the list, and his posts name the people he quotes from the firms on it.

For a Head of Marketing, this is the gap to show the partners. A firm can hold an AFR rank and still be missing from the answer that prospects read first.

Why Do Rankings Beat Service Pages?

Rankings articles beat service pages because a "top firms" query asks for a list. Wix Studio's AI Search Lab analysed 75,000 AI answers and found list articles took 40.86% of citations on commercial questions, where buyers compare options. Product pages took 7.14%, and homepages took 1.70%.

A service page answers "what does this firm do". It does not answer "which firms lead the market", so the engine looks elsewhere. That is a format mismatch, and it has nothing to do with the quality of the firm's advice.

The same firm can win and lose in one afternoon. Ask an engine for an R&D tax accountant in Melbourne, and a good service page fits. Ask for the top firms in Australia, and the engine needs a ranked list instead.

The same Wix study looked at the top 1,000 professional services URLs, where third-party listicles took 80.9% of listicle citations.

Lists that firms wrote about themselves took the other 19.1%, so the engine prefers a list someone else wrote.

This is where two gates come in. Each one is decided by something different.

Our guide on why categorisation beats authority explains the second gate in more detail.

Which Sources Can a Mid-Tier Firm Enter?

A mid-tier firm can get into all 4 source types in that answer. It fully controls only 1 of them, its own ranking post. It can earn quotes in journalist posts and trade stories, and ask list sites to fix its details. The AFR list itself ranks on revenue, so no page changes a firm's place on it.

Journalist posts

Tadros's posts name the people he quotes. A firm earns a line by giving the reporter a clear view on the list, the market or a trend such as AI adoption. That means a named partner with a view ready before the list drops each November. Keep a 2-line comment on the new list on file, with the partner's name and title.

Trade publications

Trade sites such as accountantsdaily.com.au and consultancy.com.au sat in the same answer. A firm enters them by giving an editor facts they can print. That means the new rank, the reason for the move and a named person to quote.

Directory and list sites

List sites such as profitbooks.net also appeared in the answer. A firm cannot write these pages, but it can check that its name, offices and specialisms are right and ask for fixes. A wrong office count on a list site becomes a wrong fact in the answer.

The firm's own ranking commentary

This is the only source a firm fully controls, and it already appears in the answer. On 19 August 2026, Kelly+Partners' post about its AFR rank sat 8th in Google's normal results for this search. Hall Chadwick's post sat 4th for a similar search.

Both firms turned a list result into a page of their own, such as Kelly+Partners' post on ranking 17th.

This is also why expertise alone stays invisible. Our piece on why deep expertise looks invisible shows the same pattern across law, accounting and engineering.

What Should Firms Publish When the List Drops?

A firm should publish a ranking post that answers the question in its first lines. Kelly+Partners' 25 November 2025 post opens with its rank, its movement and its growth. It ranked 17th, up 2 places, with 30% average annual revenue growth since 2006. An engine can lift those facts without reading further.

Position on the page matters as much as the facts. CXL's study of 100 citations in Google AI Overviews was published in March 2026. It found 55% of the quoted text came from the top 30% of the page, so a rank that sits below 3 paragraphs of thanks sits where fewer citations come from.

A ranking post built as an answer covers 5 things:

• The rank and the movement. State the number and the change from last year in the first sentence.
• The revenue basis. Say the AFR ranks on revenue, so readers know what the number measures.
• The specialisms. Name the 2 or 3 services and industries the firm leads in.
• The locations. Give the office count and the states, since list sites repeat these facts.
• A named partner's view. Add 1 quote on the market, which gives journalists a line to reuse.

Leave the thank-you notes, team photos and award badges lower on the page. They matter to staff and clients, but they push the rank and movement below the part engines quote most.

Our guide to getting cited in Google AI Overviews covers the page rules behind this in more detail.

How Do You Run the 30-Second Check?

You check your firm's presence by searching "top accounting firms australia" and reading the AI Overview. First, look for your firm's name in the answer text. Then open the sources and note which types filled it. The whole test takes about 30 seconds and runs in front of a partner.

Score the result on 3 levels:

The partial result is a real risk for any firm that publishes its own list. Lily Ray's study of Google AI Overviews shows how often it happens:

• 100 queries tracked. She checked business-to-business (B2B) "best category" queries between April and June 2026.
• 80 answers returned. Those queries triggered an AI Overview.
• 323 own-list citations. Brands' own listicles were cited as sources 323 times.
• 224 left out. The brand was missing from the recommendation 224 times, which is 69% of those citations.

Treat one run as a snapshot, not a measurement. Run the same query again a week later and on a second engine such as ChatGPT.

Record 4 things each time you run it:

• The date and engine. Answers change, so each result needs a timestamp.
• The firms named. List every firm in the answer text, in order.
• The source types. Tag each source as journalist, trade, list site or firm post.
• Your result. Note pass, partial or fail against the table above.

Our generative engine optimisation service tracks the same answers on each engine over time.

What Should Marketing Managers Do Now?

A marketing manager should run the 30-second check this week and fix one source type before the next AFR list. Start with the firm's own ranking commentary, because it is the only source the firm fully controls. Then brief a named partner for journalist and trade enquiries in November.

The plan fits into 3 steps:

1. This week. Run the 30-second check and note which source types filled the answer.
2. Before November. Rebuild the firm's last ranking post so the rank and movement sit in the first lines.
3. When the list drops. Brief a named partner for journalist and trade enquiries.

The work is small but time-bound. The AFR list lands once a year, and the posts written that week feed the answers buyers read for 12 months, so a firm that misses the window waits a year for the next one.

Clients already ask AI first. In a 2025 Ravical survey, 500 senior decision-makers at UK accountancy firms reported that 64% of businesses consult ChatGPT first, before they call their accountant.

Want to know where your firm stands before November? Intelligent Resourcing maps the answers your buyers read and shows which sources your firm is missing. See how answer engine optimisation works.