Sydney, Australia, Aug. 03, 2026 (GLOBE NEWSWIRE) — Somantra, an answer engine optimisation platform, has published “AI Search Ranking Factors for Insurance Brands: The Australian Citation Report,” an analysis of 2,437,107 AI search citation records covering how insurance brands are cited across ChatGPT and Google.
The report’s central finding is that ChatGPT changed how it cites insurance brands during the second quarter of 2026. In July, brand homepages accounted for 12.8 percent of all ChatGPT citations observed in the Australian insurance category, up from 1.4 percent in March. Across the seven observed months, homepages represented 3.0 percent of ChatGPT citations against 1.6 percent of Google citations.
The study finds that this shift moves brands into a weaker position rather than a stronger one. In the observed data, NRMA recorded 487 homepage citations, Budget Direct 466 at an average position of 4.80, AAMI 440 at 4.51, Allianz 426 at 4.54, and Youi 389 at 6.10. Specific answer pages ranked materially higher. Moneysmart’s pet insurance page recorded 1,306 citations at an average position of 2.23, its contents insurance page 864 at 2.05, and its total and permanent disability page 853 at 1.93.
“A homepage citation is brand awareness. A specific answer page citation is authority,” said Arun Prasad of Somantra. “What the data shows is that ChatGPT has built an associative map between brands and their products, so it no longer needs a product page to answer the question. It falls back to the homepage, and the brand loses the position it would have held with a page that answered the question directly.”
Of the 28,725 unique domains in the dataset, 5,522 appeared on both platforms. A further 18,279 appeared only in Google results and 4,924 only in ChatGPT.
The study measures a platform parity ratio for each brand, expressed as ChatGPT citations divided by Google citations. Moneysmart records the highest parity in the dataset at 29.6 percent, followed by Finder at 27.2 percent and Canstar at 21.0 percent. Among insurers, AAMI records 17.4 percent, Suncorp 14.5 percent, Youi 14.3 percent, NRMA 12.2 percent, Allianz 10.9 percent and Budget Direct 9.7 percent. Comparethemarket records 5.1 percent.
Several publishers outside the category are cited far more heavily in ChatGPT than their Google presence would suggest. Kiplinger appears at 28.3 times its Google citation count, the Wall Street Journal at 20.1 times and Wikipedia at 5.6 times.
Of the domains tracked across the seven observed months, 57.2 percent were cited in a single month and never returned. Only 279 domains, or 2.7 percent, appeared in all seven.
Content format predicts which side of that line a domain falls on. Pages using discount or savings language appear twice as often among domains surviving all seven months as among domains cited once. Comparison content survives at 1.9 times and how-to content at 1.5 times. Guide format content runs in the opposite direction, appearing in 2.0 percent of surviving domains against 7.3 percent of single-month domains.
The study also identifies formats where consumer demand outruns published supply. Coverage question pages of the form “does X cover Y” attracted 10,120 ChatGPT citations while accounting for 1.8 percent of published content. Comparison sites earn 19.2 percent of their citations from comparison format pages. Insurers earn 2.4 percent of theirs.
Transactional pages underperform in the same dataset. Quote, buy and apply language appears in 6.7 percent of ChatGPT citations against 12.4 percent of Google citations, and 91 percent of ChatGPT citations in the category serve informational queries.
“Insurers have spent a decade building pages that convert. ChatGPT is not reading those pages,” said Yajat S, the report’s author and a researcher at Somantra. “The category has an answer layer missing above the conversion layer, and the formats that fill it are cheap to publish.”
“AI Search Ranking Factors for Insurance Brands: The Australian Citation Report” analyses 2,437,107 citation records covering 28,725 unique domains across ChatGPT standard search and Google, drawn from Somantra’s generative engine optimisation index. Data covers seven observed months between November 2025 and July 2026: November, December, January, February, March, May and July. Each record carries month, platform, domain, URL, position, title and description. ChatGPT accounted for 346,172 records at an average position of 3.93 and Google for 2,090,935 at an average position of 5.77. Monthly ChatGPT sample sizes vary with the observed query set, and the report advises normalising monthly comparisons rather than reading raw counts as brand momentum.
The full report, including category level citation share, URL slug performance data, platform parity tables for every tracked brand and a content format checklist, is available at https://somantra.ai/
About Somantra
Somantra Pty Ltd tracks and measures how brands appear in AI search platforms including ChatGPT and Google. The company works with B2B and consumer brands across financial services, insurance and other regulated industries to monitor AI search citations and build content strategies for answer engines.
- AI Search Ranking Factors for Insurance Brands: The Australian Citation Report