• AI visibility

Guide · AI search

How can an insurance carrier or agency win more quotes when shoppers ask AI?

By being named, and described correctly for each state and line of coverage, when shoppers ask AI assistants to explain and compare policies. Nearly a third of auto insurance shoppers already use AI tools, and those who do switch insurers more often. For carriers and agencies, that makes AI answers a new place where quote requests are won or lost before anyone visits a comparison site.

By Underneath Team Guide Updated

The short version

  1. AI is already in the shopping journey: J.D. Power (opens in a new tab) reports 32% of auto insurance shoppers used AI tools, and they were more than 1.3 times as likely to switch insurers.
  2. Shoppers compare more than ever: J.D. Power’s 2026 shopping study (opens in a new tab) found 53% of customers shopped, collecting 3.5 quotes on average, and 48% of new auto policies were bought digitally.
  3. Each switch carries real premium: among switchers, the median premium moving between carriers is more than $3,200, according to J.D. Power’s quarterly shopping report.
  4. Life insurance needs explaining: LIMRA and Life Happens estimate 92 million US adults need coverage or more of it, and only 32% call themselves very or extremely knowledgeable about it.
  5. Google answers most insurance searches with AI: in our study of 1,248 searches, 88.0% of financial services and insurance keywords showed an AI Overview.

A note before you read: this article is about how insurers appear in AI answers. It recommends no policy or carrier, and it is not legal or regulatory advice.

Who shops for insurance now, and what is a new policyholder worth?

Households at renewal or after a life event, comparing several quotes, and increasingly buying online or after online research.

Auto insurance is the front door. In the J.D. Power 2026 U.S. Insurance Shopping Study of 12,437 customers, the share shopping fell from 57% to 53% as rate increases cooled, but customers collected an average of 3.5 quotes, the most in the study’s history. Nearly half (48%) of new auto policies were bought digitally, up from 36% five years earlier.

A new policyholder is worth more than one policy. J.D. Power’s quarterly report puts the median premium moving between carriers among switchers at more than $3,200. The bigger prize is the household: 45% of active auto shoppers said they have a homeowners policy, but only 20% received a homeowners quote while shopping. As J.D. Power’s Stephen Crewdson put it, if the auto quote isn’t competitive, customers “don’t stick around to discuss home, life or other financial products.”

Life insurance follows a different path. The 2026 Insurance Barometer, a survey of 5,237 adults, found 52% own at least one policy and 38% have a need-gap. 43% would research online but buy in person, 25% would research and buy entirely online, and 45% work with a financial professional. Demand is growing: LIMRA (opens in a new tab) reports new annualized individual life premium reached $12.7 billion in the first nine months of 2025, a 12% increase.

For agencies, this means the research often happens before the first call. For carriers, it means the comparison starts before the quote form. Health plans run on their own enrollment calendar; see how health insurers win members through AI.

Where is AI already part of insurance shopping?

At the explanation and comparison stage: shoppers use AI to understand coverage, compare policies and decide whether to switch.

J.D. Power’s data is the most direct evidence:

  • Use. 32% of auto insurance shoppers used AI tools during their search, most often for general questions, quotes, policy comparisons and decisions.
  • Quality. A similar share, 33%, found the content unhelpful.
  • Switching. Shoppers who used AI were more than 1.3 times as likely to switch insurers as those who did not. Newer digital insurers court these switchers, as how insurtechs win switching shoppers explains.
  • Understanding. Only 58% of customers said they completely understand their auto policy, down 4 points from 2025. J.D. Power’s reading is that when insurers fail to explain coverage, customers turn to AI to fill the gap.

J.D. Power’s quarterly report adds an early look at a new study: two-thirds of consumers use AI as part of researching insurance coverage, and more than one-third made a policy change based on the advice.

Google adds AI to many of these searches by default. Across the eight industries in our AI Overviews study, financial services and insurance had one of the highest rates: 88.0% of keywords showed an AI Overview, or 77.7% after adjusting for the mix of searches.

Our inference: in insurance, AI is acting as the explainer that many carriers’ own materials are not. Whoever supplies the clearest explanation has a good chance of shaping the shortlist.

Which questions do insurance shoppers ask AI?

Questions about coverage, cost, comparisons and claims, almost always tied to a state. We wrote the shopper prompts below as examples; they were not collected from real policyholders.

LineIllustrative prompt
Auto“Cheapest full coverage for a 17-year-old driver in Texas?”
Auto“Is usage-based insurance worth it if I drive under 8,000 miles a year?”
Home“Does homeowners insurance cover water backup in Florida?”
Home and auto“Which insurers give the best bundle discount in Ohio?”
Life“How much term life insurance do I need with two kids and a mortgage?”
Agency“Independent insurance agent near me who writes flood insurance”
Claims“Which insurers handle hail claims fastest in Colorado?”

Assistants also run searches of their own to answer these. In our study of the searches AI assistants run behind the scenes, one example was “J.D. Power homeowners insurance customer satisfaction Florida 2025”, an observed search for a third-party ranking in a specific state. Ratings and rankings are part of how assistants research insurers, and the state is part of the question.

How does AI visibility turn into quotes and policies?

Through the shortlist: an answer names carriers or agencies, the shopper requests quotes, and one becomes a policy.

The path, as we infer it from the documented shopping process:

  1. Trigger. A renewal increase, a new car or home, or a life event.
  2. Explanation. The shopper asks an assistant what coverage they need and who offers it in their state.
  3. Shortlist. The answer names a few carriers, comparison sites or agencies.
  4. Quotes. The shopper requests quotes, on average 3.5, through apps, websites or an agent.
  5. Policy and household. One quote becomes a policy, renewals follow, and a competitive auto quote opens the door to home and life.

Carriers already pay heavily for the quote-request stage. EverQuote, an online insurance marketplace, reported second-quarter 2026 revenue of $195.1 million, up 25%, including $172.1 million from auto insurance and $23.0 million from home and renters, and said carriers “continue to target growth across digital channels.” That revenue comes from insurers and agents buying shopping traffic.

An AI answer sits one step earlier. Our inference: a carrier that assistants name and explain well can start conversations it would otherwise pay to buy, while a carrier that is missing must compete for the same shopper later, at a price.

What decides which insurers an AI answer names?

Independent ratings and comparison sites, the insurer’s own clear explanations, and the shopper’s location. Only some of this is documented.

Documented by platforms. Google says its systems give even more weight to strong expertise and trust signals on “Your Money or Your Life” topics (opens in a new tab), which include financial stability, and that such content “must be highly accurate.”

Observed in studies.

  • Comparison sites and insurers’ own pages are both cited. In our study of 4,051 AI Overview citations, nerdwallet.com was cited in 10.0% of the 481 AI Overviews, libertymutual.com in 7.5%, allstate.com in 6.9% and thezebra.com in 6.0%. Insurers’ own sites were among the most-cited sources in their category.
  • Ranking is not enough. In finance and insurance, 34.8% of AI Overview citations were page-one organic results for the same search, so most cited pages came from elsewhere.
  • Editorial sources lead in money questions. A study of banking questions (opens in a new tab) by Mahe Chen and colleagues found 64.6% of cited sources were editorial or third-party sites. Fund questions show the same pull toward outside sources, with assistants finding fund families in trusted fund research, as how asset managers get funds considered shows.
  • Forums matter less here. Reddit was cited in 5.7% of financial services and insurance AI Overviews in our Reddit study, among the lowest of the eight industries.
  • Assistants disagree. In our four-assistant study, recommended options for financial services and insurance questions overlapped by 0.332, against 0.543 for business software.
  • Location changes the answer. In our country study, the country gap in ChatGPT’s recommendations was 0.180 for insurance, lending and tax questions, against 0.034 for global products.

Our inference for insurers. The country study tested countries, not states, but insurance is sold and priced state by state. An insurer that is clear about where it writes each line, and what its coverage includes there, gives assistants what they need to place it in the right answer.

How do state rules shape what insurers publish for AI?

They set the limits: content that assistants repeat must be accurate for each state where it applies.

Insurance is regulated by the states, and products, rates and availability differ between them. The NAIC’s Unfair Trade Practices Act model law, the template for many state laws, prohibits misrepresentations in insurance advertising, including “any intentional misquote of premium rate,” and covers advertising on the internet.

That has two practical consequences, which we infer:

  1. Clarity by state is a compliance task and a visibility task. A page that says which lines you write in which states, and what each policy covers there, helps shoppers, regulators and assistants alike.
  2. Stale pages are a risk. Old rate examples, retired discounts or outdated availability can be picked up and repeated. Our guide to fixing wrong brand information in AI answers explains how to trace an error back to its source.

States are also writing rules for insurers’ own use of AI. The NAIC’s map of states adopting its AI model bulletin shows how quickly that is spreading. That bulletin governs decisions such as underwriting and claims rather than marketing, but it signals how closely regulators watch AI in this industry.

What does GEO look like for a carrier or an agency?

For insurers, generative engine optimization (GEO) means clear coverage explanations, consistent facts and independent evidence assistants can trust.

  1. Explain coverage in plain language, by state. What a policy covers, what it does not, and which discounts apply where. J.D. Power’s finding that only 58% fully understand their policy shows the gap.
  2. Keep facts consistent. States served, lines written, discounts, claims steps and contact details should match across your site, agency profiles and comparison listings.
  3. Earn third-party evidence. Customer satisfaction rankings, financial strength ratings, editorial reviews and comparison sites are what assistants cite. Our guide on which pages to target for AI recommendations covers ranked lists.
  4. Give agencies a local presence. For agents, accurate local profiles and reviews matter, since local answers vary more than national ones.
  5. Use video for explanation. LIMRA found YouTube (66%) has overtaken Facebook as the top platform for financial information.
  6. Review with compliance. Every page assistants might repeat is advertising under state rules.
  7. Measure by state and by engine. Answers change from run to run and between assistants; see why AI answers about your brand change.

No one can promise an assistant will recommend a particular insurer. GEO improves what assistants find, so that when they do mention you, the description is correct.

What can’t the research tell insurers yet?

How assistants choose among insurers state by state, and how many policies AI answers produce.

  • Surveys, not logs. The AI usage figures come from J.D. Power surveys, which show behavior shoppers report, not what assistants said to them.
  • Countries, not states. Our location study compared countries. We found no public study of how AI answers differ between US states.
  • No link to bound policies. We found no public data connecting an insurer’s AI visibility with quotes or policies written.
  • Fast change. Assistants and Google’s AI features change often, so any snapshot ages quickly.

Where should an insurance company start?

Ask assistants your customers’ coverage and comparison questions, state by state, and record who is named and why.

That check shows whether assistants know where you write each line, whether your coverage and discounts are described correctly, and which ratings, comparison sites or competitors shape the answers instead of you. For agencies, it shows whether you appear for local questions at all.

If more quote starts, bundles or life applications are on your plan, talk to us about an AI visibility review for your lines and states. We will test the questions behind your most valuable lines in the states that matter, trace the sources behind each answer, and plan the content, coverage and data fixes, reviewed with your compliance team, that give you a fair chance of being named. How carriers and agencies keep coverage explanations and state-by-state facts consistent over time is described on our generative engine optimization service page.

Frequently asked questions

Do AI assistants give insurance quotes?

Some shoppers use AI to compare quotes, and J.D. Power counts quotes among common uses. A binding price still comes from the insurer or agent, so assistants mostly shape who gets asked.

Why does an assistant name comparison sites instead of insurers?

Comparison sites publish the rankings and explanations assistants draw on. In our AI Overview study, NerdWallet and The Zebra were cited often, but so were Liberty Mutual and Allstate’s own sites.

Does AI search matter for independent agencies?

Yes, especially for local questions. Shoppers ask for agents near them, and answers that name a place agree far less across assistants, so accurate local profiles and reviews matter.

Can an insurer correct a wrong AI answer about its coverage?

Usually by correcting the source. Find the page the answer relies on, often an old page of your own or a third-party listing, and update it.

Sources

Free strategy call

Some questions are easier to answer about your own business.

Bring the one that matters most. On a free 30-minute call we’ll take a first look at it and send you a short written read afterward.

Ask about my businessHow we work