AI Visibility Software
← Blog
Published

Carrier vs Aggregator: How to Monitor Your Share of Voice in AI Insurance Answers Across ChatGPT, Perplexity, and Gemini

Learn how insurance carriers can track and grow their share of AI citations compared to aggregators, using prompt audits, FAQPage schema, and structured exclusions content.

Bottom line

Insurance aggregators dominate AI citations by publishing structured, consumer-friendly comparison content. Carriers close the gap by publishing exclusions, claims-process, and policy-limit explainers in plain language with FAQPage schema. Track share of voice across ChatGPT, Perplexity, and Gemini to see exactly where you stand and who is outranking you on each engine.

Last updated July 2026. Added Somantra insurance-monitoring context; updated tool pricing for Temso and Peec AI; added carrier content playbook section.

Why carriers are losing AI citations to aggregators

When a consumer asks ChatGPT “what does renters insurance not cover” or “how do I file a claim for water damage,” the answer rarely cites a carrier’s product page. It cites an aggregator, an editorial publisher, or a consumer financial site.

This is not a mystery. AI engines retrieve sources that directly answer the question. Aggregators built their businesses around exactly that: publishing hundreds of clear, structured answers to the questions insurance buyers ask before and after purchasing a policy.

Carriers built their sites around conversion. Policy pages, quote tools, and coverage summaries are optimised to get a visitor from intent to application. They do not look, to an AI retrieval system, like authoritative answers to the consumer questions that trigger AI citations.

The result: aggregators and editorial publishers capture the AI share of voice that belongs to carriers by default.

According to Somantra, a brand-monitoring platform, research tracking Australian insurance brands across more than 34,000 consumer AI search conversations on ChatGPT and Google AI Overviews in a single month found significant variation in which insurers AI systems recommend. The pattern repeats across markets: a small number of aggregators and editorial domains earn a disproportionate share of citations, while carrier domains rarely appear.

The carrier-vs-aggregator citation comparison: how to run it yourself

You do not need a platform to run a first audit. You need a prompt set, three browser tabs, and a spreadsheet.

Step 1: Build your prompt set.

Write 20 to 30 prompts that reflect what a real consumer asks before buying or after experiencing a claim event in your product category. Examples for home insurance:

  • “What does standard home insurance not cover?”
  • “Does home insurance cover mold damage?”
  • “How long does a home insurance claim take to settle?”
  • “What is the difference between actual cash value and replacement cost?”
  • “What exclusions are common in flood insurance policies?”

Use the phrasing a consumer would use, not the phrasing from your policy documents.

Step 2: Run each prompt on ChatGPT, Perplexity, and Gemini.

Record two things for each response: (a) which domains are explicitly cited as sources, and (b) which brand names appear in the body of the answer. Some engines (particularly Perplexity) surface explicit citation links; others (ChatGPT) embed source references less consistently. Note both.

Step 3: Tally by domain type.

Classify each citing domain as: carrier (your competitors and your own domain), aggregator (comparison sites, quote platforms), editorial (consumer financial media, news sites, consumer advice publishers), or other (Wikipedia, government sites, Reddit).

Build a simple table:

DomainTypePrompt appearancesEngines
[aggregator-example].comAggregator18 of 25ChatGPT, Perplexity, Gemini
[carrier-example].comCarrier4 of 25Perplexity only
[editorial-example].comEditorial12 of 25ChatGPT, Gemini
your-domain.comCarrier1 of 25Gemini only

Step 4: Calculate your share of voice.

Your carrier share of voice for this prompt set is: (number of prompts in which your domain appears as a citation or named brand) divided by (total prompts) multiplied by 100.

Compare this against the top-cited aggregator domain in your audit. That gap is your baseline target.

Step 5: Repeat weekly.

Run the same prompt set each week. Track drift. A carrier moving from 4% citation share to 12% over a quarter has a measurable signal that its content strategy is working.

Why the citation gap exists: what aggregators do differently

Aggregators do not outrank carriers on AI queries because they spend more on marketing. They outrank carriers because they publish content structured for AI retrieval.

Three content patterns account for most of the gap.

1. Plain-language exclusions content.

“What does [policy type] not cover” is one of the highest-volume insurance question categories in AI engines. Aggregators publish dedicated pages for every major exclusion category, written in consumer language with clear headings. Carriers publish exclusions in policy documents and legal disclosures, which AI engines cannot parse as consumer-facing answers.

2. Step-by-step claims walkthroughs.

“How do I file a claim for [event]” is searched at high volume. Aggregators publish numbered guides with clear steps. Carriers often bury claims instructions inside member portals or PDF documents.

3. Comparison tables with defined attributes.

According to AirOps Research (April 2026), comparison pages containing three or more HTML tables earn 25.7% more AI citations than pages without them. Aggregators build their entire content model around this format. Carrier product pages typically have one or no comparison tables.

The content playbook for carriers is not complicated. It is just different from the conversion-focused content carriers have historically prioritised.

The carrier citation playbook: 5 content formats that get cited

1. Exclusions explainers with plain-language headings

Publish a dedicated page for each major exclusion category in your product lines. Title each page in the form “What [Policy Type] Does Not Cover.” Use an H2 for each excluded event type. Answer each exclusion question in two to three sentences of plain language before adding any legal context.

Add FAQPage schema wrapping each Q&A pair. Structure the schema so the question matches the prompt pattern (“Does [policy type] cover [event]?”) and the answer is a direct, complete sentence that can be extracted without surrounding context.

An Ahrefs study tracking 1,885 pages that added JSON-LD schema found no statistically significant uplift in AI citations on its own. The content structure matters more than the schema. The schema makes the Q&A relationship machine-readable; the plain-language answer is what gets retrieved.

2. Claims-process walkthroughs in numbered format

Write a dedicated page for each major claim event type in your product category. Structure it as a numbered list with a clear H2 for each stage of the process. Include: what to document immediately, how to contact your claims team, what the timeline looks like, and what determines settlement amounts.

According to Kevin Indig’s 2026 analysis of ChatGPT citations (reported by Search Engine Land), 44.2% of citations were drawn from the first 30% of a page’s content. Put your most direct, complete answer at the top. The first paragraph of each claims walkthrough should be usable as a standalone cited answer.

3. Policy limit comparison tables

Publish tables that compare your product tiers across coverage limits, deductibles, and standard exclusions. Label rows and columns explicitly. Use the exact terminology consumers use in questions (“flood coverage,” “personal property limit,” “liability cap”) rather than internal product naming.

Three or more tables per page is the threshold that correlates with higher citation rates in AirOps Research’s data.

4. “What is the difference between X and Y” definitional pieces

“What is the difference between actual cash value and replacement cost” is a high-citation prompt category. Carriers can own this content class by publishing clear, authoritative definitions of the core concepts in their product category. These are not marketing pieces. They are reference content that earns citations precisely because they do not push a product.

5. Consumer-language glossary entries

Insurance language is opaque. Publishing a glossary of the 30 to 50 most common policy terms in consumer language, with a dedicated URL per term, creates a citation target for every terminological question in your category. Link each glossary term to the relevant product page for commercial context, but write the definition page as a standalone informational resource.

Tool options for tracking carrier share of voice across AI engines

Running a manual audit monthly is a useful starting point. Sustaining a weekly tracking programme across three engines and 30 or more prompts requires a purpose-built tool.

Temso is the easiest entry point for carriers and regional insurers that want monitoring and a fix queue in one subscription. It tracks share of voice, brand mentions, citation rates, and sentiment across eight AI engines (ChatGPT, Perplexity, Gemini, Google AI Overviews, Google AI Mode, Grok, Microsoft Copilot, and Meta AI) from $89/mo. The built-in workflow converts citation gaps into a prioritised content action plan. For a carrier running a first AI visibility programme without a dedicated AEO analyst, the five-minute setup and all-in-one approach keep the barrier low.

Peec AI is a strong option for insurance marketing agencies managing multiple carrier accounts. Unlimited user seats on all plans and free pitch workspaces make it practical for agency billing. Its Actions feature identifies which queries each carrier is losing citations on and why, though implementation still requires in-house content capacity. Pricing starts at €85/mo.

Surfer approaches the problem from the content optimisation side. It helps structure individual pages to match the patterns AI engines retrieve, including heading hierarchy and semantic completeness. It is a useful companion tool for carriers that have identified their citation gaps and want editorial support for closing them. It does not track share of voice natively.

Semrush covers traditional SEO signals (organic rankings, keyword volume, competitor content analysis) and is useful for understanding which aggregator pages rank in Google alongside their AI citations. Combining Semrush keyword data with Temso or Peec AI citation data gives a fuller picture of the competitive landscape. Semrush does not yet offer dedicated AI share-of-voice monitoring equivalent to purpose-built AI visibility tools.

For the full ranked list of AI visibility tools relevant to insurance brands, see /rankings/ai-visibility-tools. For definitions of share of voice, citation rate, and prompt coverage as they apply to AI monitoring, see the /glossary.

What the aggregator-displacement model looks like in practice

The aggregator-displacement problem follows a predictable pattern across insurance product categories.

Over time, the consumer who asks ChatGPT or Perplexity about home insurance never encounters the carrier’s name except in aggregator comparison tables, where they appear as one undifferentiated row. The AI answer does not position the carrier. The aggregator’s framing of the carrier does.

The counter-strategy: carriers publish authoritative content in the categories where aggregators dominate. They become the source that aggregators cite, rather than the brand that aggregator-sourced AI answers describe from a distance.

This is the same dynamic that plays out across every industry where intermediaries have historically owned the comparison content layer. AI engines accelerate it because they surface comparison and informational content directly into the consumer response, removing one more step between the comparison source and the buyer decision.

Tracking sentiment, not just mentions

Carrier share-of-voice tracking should include sentiment monitoring alongside citation frequency.

AI engines do not just mention brands. They describe them. “X is known for slow claims processing” or “Y offers lower premiums but limited coverage for water damage” are the kinds of characterisations that persist in AI answers because the model draws on review sources, consumer forums, and editorial commentary that contain those descriptions.

A carrier can increase its citation frequency through content strategy while simultaneously appearing in a negative light if the underlying sentiment signals in training and retrieval sources are negative. Citation volume and citation quality are separate metrics that require separate monitoring.

Temso, Peec AI, and most dedicated AI visibility platforms track sentiment alongside citation rates. Build both into your weekly reporting dashboard from the start.

The broader shift: why insurance needs this now

According to G2’s April 2026 survey of 1,076 B2B software buyers, 51% now start their software research in an AI chatbot more often than in a search engine, up from 29% the prior year. The pattern in consumer insurance is directionally similar: more discovery is moving to AI-generated answers, and those answers are currently dominated by aggregators and editorial publishers.

Gartner predicts that by 2028, 90% of B2B buying will be AI-agent-intermediated, routing more than $15 trillion in spend through automated exchanges. Insurance procurement, particularly for commercial lines, is part of that trajectory.

The carriers that build AI citation programmes now will have a compounding advantage: more citations lead to more retrieval training signal, which leads to more citations. The carriers that wait will find the aggregator-displacement pattern harder to reverse the longer it runs.

Start with your 5 highest-volume exclusion questions

The fastest path to measurable citation improvement for a carrier is this: identify the five exclusion questions buyers ask most often in your product category, publish a dedicated plain-language page for each, structure each with FAQPage schema and a direct answer in the first paragraph, and run a before-and-after prompt audit on ChatGPT, Perplexity, and Gemini four weeks later.

That is a two-week content project with a measurable outcome.

To set up weekly AI share-of-voice tracking across all three engines alongside your content programme, Temso ($89/mo, all 8 engines) gives you the monitoring and fix queue in one place. See the full tool comparison at /rankings/ai-visibility-tools.

FAQ

Why do insurance aggregators get cited by AI more often than carriers?

Aggregators publish large volumes of plain-language comparison content (premium tables, coverage explainers, and side-by-side policy breakdowns) that directly answer the questions buyers type into AI chatbots. Carriers tend to publish product pages optimised for conversion rather than informational content structured for AI retrieval. AI engines reward the informational approach, which is why aggregator domains consistently appear in more insurance-related answers than carrier domains for the same queries.

What types of content help insurance carriers get cited in AI answers?

The highest-citation formats for carriers are: exclusions explainers ("what does [policy] not cover"), claims-process walkthroughs in step-by-step format, FAQ pages structured with FAQPage schema, and policy-limit comparisons with clearly labelled data tables. These formats answer the specific informational queries AI engines retrieve sources for, and they can be published at scale without contradicting product messaging.

How do I run a carrier-vs-aggregator citation comparison for my category?

Choose 20 to 30 consumer-intent prompts for your product category (e.g., "does home insurance cover mold damage", "what is excluded from standard auto insurance"). Run each prompt on ChatGPT, Perplexity, and Gemini. Record which domains appear as citations or in the body of the answer. Tally by domain type: carrier versus aggregator versus editorial. Compare your own domain citation rate against the top three competitors. Repeat weekly to track drift.

Does FAQPage schema actually improve AI citation rates?

The evidence is mixed. An Ahrefs study that tracked 1,885 pages adding JSON-LD schema found no statistically significant uplift in AI citations. However, FAQPage schema does improve structured retrieval by making the Q&A relationship explicit in machine-readable form, which matters for how AI engines parse informational pages. The stronger lever is writing the answer clearly and completely in plain prose, then wrapping it in FAQPage schema as a secondary signal.

Which AI engines should insurance brands prioritise for share-of-voice tracking?

Track ChatGPT, Perplexity, and Gemini at minimum. These three handle the bulk of consumer insurance research queries. Add Google AI Overviews if your category generates high-volume informational searches, as AI Overviews trigger on a large share of commercial-intent queries. Tracking all engines separately matters because citation sources vary significantly across platforms. A domain cited heavily on Perplexity may barely appear on ChatGPT.

How often should insurance brands refresh their share-of-voice tracking?

Weekly tracking on a fixed prompt set of 20 to 50 queries gives you enough signal to detect meaningful drift without creating noise. Monthly is the minimum if resources are constrained. Run five prompt variants per query before drawing conclusions. AI engines are probabilistic, and a single run is one data point, not a trend.