What Does a 12% Commission Actually Buy?

Jun 17, 2026

insuranceindustry.ai article hero image

AI Is About to Answer That Question.

By James W. Moore | InsuranceIndustry.AI Distribution Economics Series: Part 1

Key Takeaways

  • Your commission check has always covered four different jobs. AI is making three of them cheaper by the month.
  • Prospecting, placement on standard risks, and service are all heading toward near-zero cost. The fourth job, advocacy, is where your real value lives.
  • This is not a story about agents being replaced. It is a story about which agencies will get paid more, and which will struggle to explain what they are charging for.
  • The agencies best positioned for what is coming are already figuring out what they are actually selling. And it is not placement.

An independent commercial lines agency in the Midwest writes $8 million in annual premium. At an average commission rate of 11.5 percent, that is $920,000 a year for connecting clients to carriers.

For decades, nobody at that agency has had to formally answer what, exactly, each of those dollars is paying for. Not because the answer is obvious. Because the way the market worked made the question unnecessary.

The people who will ask it next are already building the tools to answer it themselves, and what they are finding is not good news for the current commission structure.

Commission Was Never Priced for Advice

Before we talk about what is changing, it is worth being honest about what the commission was actually paying for. It was not expertise.

Agency commission was accurate pricing for real work: figuring out which carriers would actually write a risk, building submissions by hand, chasing quotes across multiple markets, translating underwriter requirements into plain English for clients, and navigating a distribution system that was never designed to make sense from the outside.

That work was real, and so was the time it took. Commission was what the market paid to get it done.

Think about travel agents. They did not disappear because they gave bad advice. They disappeared because the hard part of their job became something anyone could do online in ten minutes: finding flights, comparing prices, booking itineraries. The advice was still valuable. But nobody was willing to pay for it once the search work went away.

Insurance is not the same as booking a flight. But the dynamic is: when the hard, time-consuming part of a job gets cheap, the price for the whole job tends to follow. The functions that hold their value are the ones that were never really about information gathering in the first place.

Commission was priced for friction. Friction is not a durable asset.

Four Jobs Inside Every Commission Dollar

Every agency does four things for every account in the book. Your commission covers all four. The problem is that AI is driving down the cost of each one at very different speeds, and three of them are moving in the same direction.

Prospecting: Finding the Risk Before It Finds Someone Else

What it costs today: a lot. Research cited by Insurance Thought Leadership put the average cost of acquiring a customer through the independent agent channel at roughly $900, compared to $487 for direct writers. The agency-side math tells the same story: producer time, marketing spend, data lists, outreach, and a sales pipeline that converts at single-digit rates on cold calls, with traditional commercial insurance leads running $100 to $300 per qualified lead.

Where it is going: a fraction of that. AI tools are already scraping permit filings, business registrations, and public data to identify prospects, score them, and sequence outreach automatically. Early AI lead qualification deployments across industries are reporting costs around $39 per qualified lead, against $260 or more for human-driven outreach. The Big ‘I’s 2026 ACT Tech Trends Report found two-thirds of independent agencies plan to increase AI use this year, with prospecting and operational efficiency at the top of the list. The cost argument for a high first-year commission on a standard small commercial account is getting harder to make every month.

Placement: Getting the Risk to the Right Market

What it costs today: This is the one most agents point to when someone asks what they actually do. Placing a complex mid-market account can take three to six hours of a skilled producer’s time: knowing which carriers have real appetite, building the submission, managing underwriter questions, and negotiating terms. For agents who have spent years learning the market, this is the work that feels most like a craft.

Where it is going: cheaper, but not evenly. Comparative raters already handle personal lines and a lot of small commercial lines with minimal human input. AI submission tools are speeding up the ACORD prep, loss run assembly, and market matching for standard mid-market accounts.

But this is also where a lot of AI commentary oversells the technology. The less standard the risk, the more placement is a negotiation, not a search. Carrier appetite for specialty, E&S, and layered programs is often undocumented, inconsistent, and built on relationships that took years to develop. An AI tool that reads the underwriting guide is not the same as a producer who knows which underwriter actually writes coastal habitational and which one just says they do. That difference matters for a wide swath of commercial business.

Standard risks are commoditizing. Complex risks are not. If you only write one type, you need to know which side of that line you are on.

Service: Everything After the Bind

What it costs today: more than most owners realize, because it hides in CSR salaries rather than showing up as a line item per transaction. COIs, endorsement requests, billing questions, renewal prep, coverage questions. Keeping a book of business running consumes the majority of service staff time in most agencies.

Where it is going: down, and it is already happening. AI handles the high-volume, routine service work at a fraction of the cost of a CSR. COI generation, endorsements, standard coverage questions: agencies that have deployed the tools are running these around the clock without adding headcount. This is not theoretical. And agencies know the rest is coming: the 2025 IVANS Connectivity Survey found 72 percent of agencies say the commercial submission process is the top area where they want more automation. They can see the runway ending.

Advocacy: The Work That Shows Up at Claim Time

What it costs today: hard to say, because most agencies have never priced it separately. It gets bundled into the commission and given away. Advocacy is what happens when the carrier’s loss run has an error that would spike the renewal, and the agent catches it before the client sees the invoice. It is what happens when the adjuster’s reading of the policy does not match what the client thought they bought. It is the pre-loss email chain the agent kept from six months before the claim, the operations change the client mentioned in passing at renewal, and the COI requirement that showed up at 8 p.m. the night before the project started. None of that is in the file. The agent carries it.

Where it is going: nowhere. This work is durable not because it is personal, but because it carries real responsibility. AI can summarize a policy. It cannot be held professionally accountable for getting the coverage interpretation wrong. It cannot use a ten-year carrier relationship to get a disputed claim reconsidered. The value here is not sentiment. It is judgment, accountability, and knowing the client’s business better than any file note ever will.

There is also a straight economic argument that rarely gets made explicitly: an agency placing $8 million with a carrier has renewal leverage that a one-off AI-generated submission does not. That leverage is real, and it is worth money. It is also the one function that most agencies have never formally charged for.

The Math Gets Uncomfortable

That is the problem in plain English. When three of the four things your commission pays for are getting dramatically cheaper, a flat percentage on a standard account gets hard to justify. Not as a moral argument, but as a business reality.

Bank of America’s 2026 research put more than $15 billion in insurance commissions, specifically low-complexity transactions across just six major carriers, in the category of directly exposed to AI-driven displacement. The mechanism is not robots replacing agents. It is carriers and platforms doing the cheap work more cheaply, and adjusting what they pay for it.

Run the numbers on that $920,000 agency. If per-account prospecting costs fall by an order of magnitude and AI absorbs most of the routine service work, somewhere in the neighborhood of $300,000 of that revenue stops corresponding to work that clients or carriers will pay a percentage commission to receive. The commission check does not disappear overnight. But the case for it on standard BOP and small commercial gets thinner every year.

There is a real opportunity in the short term, and it is worth noting: agencies that deploy AI now will see their costs drop before carriers adjust what they pay. That window, call it the efficiency dividend, is genuine and rewards early movers. But agencies that ride it too long hit a wall. Costs drop, commissions stay the same, profitability looks fine, and then the repricing arrives faster than the business model can handle. That is the compression trap, and it is what happened to travel agencies, mortgage brokers, and stock brokers before them. The window closes. Plan for it now rather than later.

Here is the other uncomfortable truth: most of the AI investment happening in independent agencies right now is going into the functions that are already cheapening. Operational efficiency, staff productivity, and prospecting automation. That is where the tools are being deployed. The advocacy layer, the one function with genuine long-term pricing power, is still largely informal, inconsistent, and free. Agencies are using AI to get faster at the least defensible parts of their value proposition, while the most defensible part goes unpriced and unnamed.

As for who comes out ahead when the dust settles, it will not be evenly distributed. Large brokers with proprietary data, AI-native MGAs with direct carrier access, embedded platforms, and elite specialty shops where advocacy is the whole product. Those are the likely winners. The agencies most at risk are the generalized mid-market shops that do a little of everything without a clear story for why a client could not get the same thing somewhere cheaper.

What to Do About It

None of the right moves are complicated. Actually making them is the hard part.

Start with the overhead. Most agencies staff based on how many transactions they handle, not on what those transactions are worth. Before the market forces a correction, take a hard look at where your CSR time is going. If your people are spending most of their day on COIs, endorsements, and billing questions, you are paying human wages for work that AI can do around the clock for a fraction of the cost. That is not a technology decision. It is a cost correction that is overdue.

Be honest about what you are actually good at. Volume placement on standard commercial accounts is moving to platforms and direct channels that can do it faster and cheaper. If that describes most of your book, you have a portfolio problem, not a technology problem. The agencies that will hold their ground are the ones with the expertise to handle accounts that require real judgment: the ones where the market is complicated, the client is hard to understand, or the claims history needs interpretation. That is a deliberate choice about what kind of agency you want to be.

Charge for the thing that is actually worth something. If your advocacy work, the claims intervention, the renewal analysis, and the coverage counseling are bundled into a commission percentage on every account, regardless of complexity, you are pricing your most valuable service like a commodity. Some agencies are already separating this: standard accounts go to a lower-touch service model at a reduced rate, while complex accounts with real advocacy needs carry a separate retainer or fee. Agencies that have tried this report that revenue per account holds, and that clients who understand what they are paying for tend to stick around. One important caveat before you restructure anything: the rules around charging fees in addition to commission vary significantly by state. Some states allow it freely with written disclosure. Others restrict it to specific lines or account types. A few effectively prohibit it outside of narrow circumstances. The NAIC publishes a state-by-state chart of producer fee and commission rules worth reviewing before you change how you charge. If you cannot explain what your advocacy is worth, you cannot charge for it. Start there. Then check your state’s rules.

The Commission You Can Actually Defend

Commission made sense for a long time because the work it covered was hard, time-consuming, and genuinely valuable. That is still true for part of what your agency does.

The part that is getting cheaper is the part that was always the most visible: finding the client, getting the quote, issuing the certificate. The part that holds its value is the part that most agencies have never had to explain out loud, because nobody asked.

The agencies that figure out how to explain it and charge for it will be fine.

The ones who cannot answer the question will not be replaced.

They will just be repriced.

 

Sources

 

Subscribe to AI Insights Newsletter

AI Disclaimer: This content was created with assistance from artificial intelligence technology. While content is based on factual information from the source material, readers should verify all details directly with the respective sources before making business decisions.

Related Posts

468