Insurtechs spend the entire budget on the policyholder and almost nothing on the person who actually decides what gets sold.

1.  Who actually decides which carrier writes a commercial risk?

A.  The independent agent does, and the business owner mostly ratifies it. In 2025, independent agencies placed 87.7% of all US commercial lines written premium, essentially flat against 87.9% the year before. Across all property and casualty, the channel placed 62% of a market that reached $1.1 trillion in direct written premium. Those are AM Best numbers via the Big “I”. Personal lines is the outlier at 39.5%, and personal lines is where nearly all the consumer-facing marketing energy has gone.

Sources: Big “I” 2026 Market Share Report (AM Best data); Gallagher Re Global InsurTech Report.

2.  So where did the innovation money actually go?

A.  About fifteen cents on the dollar. Gallagher Re has tracked $60.8 billion of cumulative global insurtech investment since 2012. Of that, roughly $9.3 billion went to commercial-focused companies. Meanwhile 99.1% of Q2 2026 funding went to firms labeled AI. The capital has been pointed at the segment with the smallest premium base and the weakest intermediary, then branded for an audience that does not place the business.

3.  Fine. But surely the broker is shopping on price like everyone else?

A.  She is not, and this is the most expensive misunderstanding in the category. Corporate Insight surveyed 214 independent agents in Q4 2025 and asked what matters most when choosing a carrier. Underwriting appetite and flexibility came first at 39%. Ease of doing business, 35%. Product breadth, 32%. Competitive pricing finished fourth at 26%. Ask personal lines agents the same question and 69% say price. Two different businesses wearing one word.

Source: Corporate Insight, Commercial Insurance Monitor, n=214, fielded Q4 2025.

4.  What does the broker’s day actually look like?

A.  Because her attention is the scarce resource, and her inbox consumes it. 61% of commercial agents report spending 60% or more of the day on administrative work rather than selling. Only 16% describe their day as tilted toward selling at all. The most time-consuming task is claims communication, at 58%; compliance paperwork and billing tie at 42%. Then the quote itself: 55% say a typical commercial quote cycle runs four to seven days, and just 7% can turn one in one to three. A carrier that removes a day of friction has done something no rate filing can do.

Source: Corporate Insight, Q4 2025 (n=214). Mid-range derived from reported extremes.

5.  How badly are carriers actually doing on this?

A.  Badly enough that it shows up in the scorecards. The J.D. Power 2025 study, now in its eighth year and built with the Big “I”, is blunt.

Measure	Figure	What it means for you
Commercial agents who say carriers meet their foundational needs	57%	Two in five say you fail at the basics
Commercial agents who feel insurers do not value them as partners	22%	One in five is already looking elsewhere
Satisfaction gap, commercial, “very easy” vs. poorly rated carriers	314 pts	Ease is the largest single lever in the study
Undervalued commercial agents likely to cut business with you	7×	The penalty for indifference is measurable
Agents rating their primary carrier’s claims service “excellent”	23%	A wide-open field for anyone who is good

Sources: J.D. Power 2025 U.S. Independent Agent Satisfaction Study (with IIABA); Vertafore agent survey, 2025.

6.  Is the broker really asking an AI assistant?

A.  She is. Liberty Mutual’s 2026 Independent Agency Growth Study surveyed nearly 1,200 agency principals and staff and found a majority used AI for work in the past year; 38% report meaningful time savings averaging four hours a week. Broader research puts adoption at 64% of agencies using AI in at least one workflow, rising to 91% at agencies with more than 25 producers. ReSource Pro puts 98% of agencies on track to invest in AI in 2026. The Big “I” Agents Council for Technology supplies the counterweight: two-thirds plan to increase use, only 8% have it in daily workflows, and 55% have no written policy. Adoption is early, messy and individual, which is exactly the phase where default answers harden into habits.

7.  When she asks, whose pages does the machine actually read?

A.  Almost never yours. AirOps’ 2026 State of AI Search report found 85% of brand mentions in AI answers originate on third-party pages rather than owned domains, making brands 6.5× likelier to be cited through someone else’s site. Stacker and Scrunch ran the controlled version: the same article on a brand’s own site was cited 8% of the time; distributed through third-party news outlets, 34%, a 325% lift, with a 239% median confirmed on a broader sample in March 2026. And 80% of cited URLs do not rank in Google’s top 100 at all, so your SEO scoreboard is measuring a different game.

Sources: Stacker/Scrunch controlled study, Dec. 2025 (944 prompt-platform combinations); AirOps 2026 State of AI Search.

8.  Airlines already did this to travel agents. Settled?

A.  It settles the opposite of what people think. The collapse was real: US retail travel agency locations fell from 22,938 in 1997 to 9,387 in 2013, a 59% wipeout, and employment sits about 60% below its 2000 peak of roughly 340,000. But look at what happened next. The survivors moved upmarket into complex, high-value bookings. Travel agent unemployment in 2025 was 2.8%, lower than it was before the dot-com boom. Median pay: $50,160, and BLS now projects essentially no change in employment through 2035 with about 5,900 openings a year. What died was typing a point-to-point flight into Sabre. What lived was judgment. Commercial insurance has no point-to-point flights.

Sources: Skift, 1997–2013; Stripe Economics, April 2026; US BLS Occupational Outlook Handbook, May 2025.

9.  Celebrities, they’re just like you. Ask Taylor Swift.

A.  That fame buys you no distribution. In November 2022 the most powerful recording artist alive put Eras Tour tickets on sale. 3.5 million people registered for the presale. The system absorbed 3.5 billion requests in a day, four times its old record, and fell over. Ticketmaster sold 2.4 million tickets, an all-time single-day record for an artist, then canceled the general sale outright. The tour grossed about $2 billion. None of that let her route around the intermediary. It took a Senate Judiciary hearing in January 2023 and a DOJ antitrust suit to even start the argument. If Taylor Swift cannot buy her way past the channel that holds the transaction, your Series C cannot either.

A.  The corporate version is Zillow. It had the traffic, the brand and the algorithm, and in 2018 chose to buy and flip houses instead of serving agents. In November 2021 it wound down Zillow Offers, took a write-down above $540 million and cut 2,000 jobs, about 25% of the company. Here is the part that should keep a CMO up at night: that same quarter, Premier Agent, the business that sells to agents, booked $358.9 million and grew 20% year over year. The disintermediation play lost half a billion dollars. The serve-the-channel play paid the bills. Zillow still runs Premier Agent today.

10.  And the insurtechs that spent it all on the consumer?

A.  The unit economics were visible in the S-1. Lemonade spent $0.76 of sales and marketing for every $1 of new gross written premium. Allstate, working through a distribution channel, spent about $0.26. In 2020, Lemonade put $80.4 million into sales and marketing, 37.4% of total expenses, more than it spent on losses. Cumulative losses since inception passed $1.27 billion by the end of 2024. Hippo’s CFO put its 2020 loss ratio at 121%. All three marquee direct insurtechs now distribute through the independent agency channel. Hippo’s own site tells agents it invests to empower them. Call that arithmetic catching up with strategy.

11.  What do I publish, concretely, starting Monday?

A.  Appetite. Plainly, publicly, in specifics a machine can lift. A third of commercial submissions end in “not a fit” after an agent has already spent twenty to thirty minutes keying ACORD data. There are roughly 2,700 US P&C carriers; the largest API-based rater reaches about 48 of them, and just about 131 offer an online quoting portal an independent agent can actually use. The retrieval layer is starving for the information you are hiding in a PDF behind an agent login.

Page to build	What goes on it	Why it gets retrieved
Appetite statement	Class codes, revenue and TIV bands, states, and the classes you decline	Specific, factual, unhedged: the highest-citation content type at 38–65%
Service standards	Named quote turnaround in hours, submission-to-decision, endorsement SLA	A number is quotable; “industry-leading service” is not
A hard claim, told properly	One real claim, with permission, dates and dollars intact	Brokers trade claims stories, not rate stories
Appointment mechanics	Production minimums, commission structure, onboarding timeline	Removes the reason she never called you

Citation-rate ranges: Clearscope and AirOps, 2026. Carrier counts: QuoteSweep market research, 2026.

12.  How do I find out whether I am invisible?

A.  Run the audit in her vocabulary. Put these into the assistants your agents use and see whether your name appears at all. Note that 44.2% of LLM citations come from the first 30% of a page’s text, so the answer belongs in your first forty to sixty words, not paragraph nine.

What the broker types	What you should be able to prove
“who writes habitational in Texas over $10M TIV”	You appear, by name, with the actual band
“fastest carrier for small contractor GL quotes”	You have published a turnaround number
“carriers that will write a restaurant with liquor over 40%”	Your declines are as public as your appetites
“is [your company] good to work with as an agent”	Something other than your homepage answers this

The Bottom Line

You cannot buy your way past a distribution channel that holds the relationship, the trust and the renewal. Travel agents did not die because the internet arrived; the ones who only typed did. The ones who handled complexity are at 2.8% unemployment. Commercial insurance is nothing but complexity, which is why 87.7% of it still moves through a human being who is, at this moment, typing a question into a chat window.

The good news is embarrassing in its simplicity. The agent is asking a question almost no one has bothered to answer in public: what will you write, how fast will you answer, and what happens on a hard claim at 4pm on a Friday. Publishing that is cheaper than a Super Bowl spot and it is the exact format the retrieval layer rewards. Being easy to place is a marketing problem, not an underwriting one, and right now it is a marketing problem with almost no competition.

Sources

1. Big “I” 2026 Market Share Report (AM Best data): 87.7% commercial, 62% all P&C, 39.5% personal, $1.1T DWP. https://www.independentagent.com/news/big-i-releases-2026-market-share-report/

2. Gallagher Re, Global InsurTech Report Q3 2025 / Q1–Q2 2026: $60.8B cumulative, $9.3B commercial, 95.2% and 99.1% AI share. https://www.ajg.com/gallagherre/news-and-insights/global-insurtech-report-for-q1-2026/

3. Corporate Insight, Commercial Insurance Monitor (n=214, Q4 2025): Appetite 39%, EODB 35%, price 26%; 61% admin; 55% four-to-seven-day quotes. https://corporateinsight.com/what-independent-agents-really-want-from-commercial-lines-carriers/

4. J.D. Power 2025 U.S. Independent Agent Satisfaction Study: 57% foundational needs, 22% undervalued, 314-point gap, 7× penalty. https://www.jdpower.com/business/press-releases/2025-us-independent-agent-satisfaction-study/

5. Vertafore agent survey via Carrier Management, Feb. 2025: 84% underwriting responsiveness, 78% communication, 23% excellent claims. https://www.carriermanagement.com/news/2025/02/25/272159.htm

6. Liberty Mutual 2026 Independent Agency Growth Study (n≈1,200): Majority using AI; 38% saving ~4 hours per week. https://www.agentforthefuture.com/topics/technology/artificial-intelligence-for-insurance-agents/ai-adoption-in-2026/

7. Big “I” Agents Council for Technology, Tech Trends 2026: Two-thirds increasing AI use; 8% daily; 55% no written policy. https://www.insurancebusinessmag.com/us/news/technology/clients-now-expect-brokers-to-lead-on-ai-not-just-advice-583895.aspx

8. Stacker / Scrunch controlled citation study, Dec. 2025 & Mar. 2026: 8% owned vs. 34% third-party citation rate; +325%, +239% median. https://authoritytech.io/blog/machine-relations-evidence-earned-media-ai-citations

9. AirOps 2026 State of AI Search: 85% of brand mentions third-party; 6.5× citation likelihood. https://www.airops.com/blog/first-party-third-party-citations-llms

10. Skift / Stripe Economics / US BLS: 22,938→9,387 agencies; –60% employment; 2.8% unemployment; $50,160 median. https://www.bls.gov/ooh/sales/travel-agents.htm

11. Zillow Group Q3 2021 Form 8-K and CNN, Nov. 2021: $540M+ write-down, 2,000 jobs, Premier Agent $358.9M same quarter. https://www.sec.gov/Archives/edgar/data/1617640/000161764021000085/q32021991.htm

12. Taylor Swift–Ticketmaster, Senate Judiciary hearing Jan. 2023: 3.5M registered, 3.5B requests, 2.4M tickets, general sale canceled. https://www.nbcnews.com/politics/congress/live-blog/ticketmaster-taylor-swift-senate-committee-hearing-live-updates-rcna66687

13. Lemonade S-1 analysis; Insurance Thought Leadership, Mar. 2025: $0.76 vs. $0.26 per $1 GWP; $80.4M S&M; $1.27B cumulative losses. https://www.insurancethoughtleadership.com/telematics/lemonade-hippo-and-root-are-back

14. QuoteSweep market research, 2026: ~2,700 carriers, ~48 API-reachable, 131 with agent portals, one-third “not a fit”. https://www.quotesweep.com/blog/commercial-insurance-quoting-tools-compared