Budget Direct took ChatGPT's insurance ad slot almost alone: 7,029 ads while Allianz, NRMA and AAMI ran zero
Somantra, a Sydney-based AI search brand visibility platform, ran 34,279 unbranded insurance queries through ChatGPT from Australia and found one company buying the ad slot almost alone: Budget Direct, with 7,029 ads — 60.7% of every ad that actually sold insurance. The rivals ChatGPT names most often bought nothing at all.
One Australian insurer is buying ChatGPT's insurance ad slot almost alone. Research released on Monday by Somantra, a Sydney-based AI search brand visibility platform, found that Budget Direct was the only one of 20 tracked insurers with ads selling insurance across 34,279 unbranded ChatGPT insurance queries. (EINPresswire, Somantra)
The numbers are stark. Budget Direct recorded 7,029 ad appearances — 19.9% of all ads shown, and 60.7% of the ads that sold insurance. Its ad appeared beneath 19.6% of all responses. Together with the broker Consolidated Insurances, Budget Direct accounted for 92% of the insurance sector's ads.
The empty ad slot
The insurers ChatGPT mentions most — Allianz, NRMA and AAMI — had no ads at all, according to the study. Neither did Canstar, the name ChatGPT mentioned and cited most often across the whole sample. RACV recorded five ads, but all of them promoted resorts rather than insurance. Brokers and comparison sites held 32% of the ads that sold insurance.

A focused conquest
Budget Direct did not just dominate the slot — it targeted it. The company's ad appeared beneath 21% to 35% of the responses that mention each of its eight most-mentioned rivals. Somantra found the campaign was tightly focused: 39 ad versions linking to just five pages, all for car insurance, with the most common copy built around "online quote in minutes" appearing on 43.8% of its ads.
The reach math, per the study: being included in ChatGPT's answers puts Budget Direct in front of 1 in 10 people asked. With its ad presence added, that rises to more than 1 in 4.

"Two separate contests"
"The insurers ChatGPT names most often have left the ad slot to a single competitor," said Arun Prasad, founder of Somantra. "An insurer can appear in thousands of answers and still have a rival's ad directly underneath. Being mentioned and being advertised are now two separate contests. Most Australian insurers are competing in only one, and many are not yet measuring either."
How the study was run — and its limits
Somantra asked ChatGPT 34,279 insurance queries from Australia on desktop, as a user with no history, in August 2026. No query named an insurer, so these were unbranded queries. The advertiser analysis reflects two days of collection, 1 August and 3 August 2026 — an insurer with no ads on those days may advertise on other days. The study counts recorded ad appearances only: it does not observe spend, bids, clicks or sales.
Why it matters
ChatGPT ads are a new distribution layer, and this is the first hard look at who is actually buying it in one national market. The pattern is familiar from search-engine history — the organic answer and the paid slot are separate auctions — but the speed matters: rivals were named in thousands of answers while a competitor's pitch sat directly underneath. For brands investing in AI search visibility (AEO/GEO), the takeaway is that showing up in the answer is only half the contest. The full report is available free from Somantra. (Somantra)