Research from Somantra found that roughly seven in ten ChatGPT responses to insurance queries contained at least one advertisement, yet the placement of those ads was highly concentrated — 327 of the 371 identified advertisers together accounted for only 11% of all ad slots.
A handful of advertisers — just five in total — held a combined share of more than half of all insurance-related ad placements appearing within ChatGPT search results, their share reaching 57 percent. Leading the field by ad volume, Budget Direct — an insurer — placed 7,029 advertisements, accounting for 19.9% of the total; Jotform, a form-building platform, followed with 5,285 ads at 15.0%; Variant Group's pdf.net tool contributed 3,372 ads representing 9.6%; Consolidated Insurances, operating as a brokerage, recorded 2,611 ads at 7.4%; and comparison platform Finder rounded out the group with 1,791 ads, equal to 5.1% of all placements.
A further 39 advertisers accounted for 31.9% of all ads, each placing somewhere between 100 and 999, forming a mid-tier cluster that tapered off well below the top of the market. Of the full advertiser pool, 327 names — accounting for 88% of those listed — collectively held just over one-ninth of the total ad inventory. Among that total, 89 turned up only a single time.
Advertising placements have become a routine feature within the results page. Across the dataset, advertisements outnumbered replies at a rate of 103 to every 100, with just under half of all replies carrying a single ad, roughly a quarter bearing two, and about one in twenty-five holding three. Of the responses examined, the maximum tally for any single entry stood at three, while just over a quarter — 28.8% — contained none at all.
Beyond that, 55 advertisers of smaller scale named a city or state, ranging from removalists in Sydney and security firms in Perth through to regional car dealers and tradespeople, and a further 26 operated from abroad, among them insurance providers based in the United States and Canada.
Data from the home and property category showed off-topic rates of 33.2% among smaller advertisers, nearly double the 18.2% recorded for larger ones. Arun Prasad, founder of Somantra, characterised the insurance ad inventory within ChatGPT as resembling a consolidated market — dominated by a handful of heavy buyers, trailed by numerous smaller ones, with only one of those five leading purchasers actually being an insurer.
Across the Australian insurers tracked, those outside the front-runner have largely left that ground to be divided among form builders, PDF tools, brokers, and comparison platforms, with no single provider stepping in to fill it cohesively. What an advertiser actually secures through a purchase is a specific location on the page, nothing more.
Brands that invest heavily in paid placements can achieve a page presence far exceeding what organic search alone delivers — Budget Direct, whose name surfaced organically in 10.0% of ChatGPT responses, appeared in 26.7% of cases once purchased placements were counted, a 16.7-percentage-point difference that illustrates how advertising spend can broaden a brand's footprint on the platform beyond what unprompted mentions would produce on their own.
On 1 August and 3 August 2026, a total of 34,279 insurance queries originating from Australia were put to ChatGPT via desktop by Somantra, with the account carrying no prior interaction history at the time of each session. None of the queries run through the system carried any insurer's name or brand identifier.
Each time an ad appears in a response, that instance registers as a separate occurrence, meaning ten responses displaying the same ad produce a tally of ten distinct appearances. Advertiser placement within the rankings reflects ad visibility alone, with spending levels, bid amounts, click data, and transaction figures all falling outside what Somantra tracks, meaning the ads any individual user encounters may vary from those shown here.
