AI-native brokerages bring in nearly $1 million per employee in Zywave report
A new Zywave study puts AI-native brokerages at close to $1 million in revenue per employee, versus about $175,000 at traditional agencies. Zywave’s data points to the back office as a key source of the gap, even as producer roles grow. It also argues cleaner, automated data can affect carrier bonus commissions, which run from 0.43% to 6.74% of premium on one published schedule.
This story was produced through MarketScale. See how Business Services teams put it to work with Executive Thought Leadership.
Key takeaways
When a carrier bonus can swing fifteenfold on loss ratio alone, the data work that trims an agency’s back office can also influence the size of its contingent or bonus check—especially if it helps keep the book in better loss-ratio tiers.
Revenue per employee is the planning number to track: about $175,000 at traditional agencies and close to $1 million at AI-native brokerages, while P&C headcount grows by less than 1% a year.
Survey adoption overstates readiness. Nearly three-quarters of brokerages report some AI use, but only 13% use it across multiple parts of the business.
Free workspace
Turn your Business Services expertise into content.
Record interviews, organize footage, and write with AI on a free trial of the MarketScale platform for qualifying companies. No demo required, no credit card.
A traditional US insurance agency brings in roughly $175,000 for every person on its payroll. An AI-native brokerage brings in close to $1 million, nearly six times as much. Zywave, the Milwaukee-based insurance technology provider, documents that gap in a new study, Insurance Business America reports. The figure is not a projection. Economics like these are already splitting the US distribution market in two.
The Milwaukee insurance technology provider built the comparison from employment data, commission structures and technology adoption across US distribution. Its main conclusion is milder than the ratio. AI is not taking the broker's seat. Firms that have automated their back offices are growing revenue faster than they add staff, though, and the distance between them and everyone else can already be measured.
Picture an agency principal deciding whether to hire the next account manager or buy the next workflow tool. The useful question is which part of the business the gap actually sits in. Zywave's data points to the back office. It also points to a second payoff few agencies connect to automation at all: the carrier bonus check.
Paper jobs shrink, client jobs grow
More than half of agency staff, 55.6% by Zywave's count, still work in service and back-office roles. That means chasing documents, re-keying data and handling routine policy servicing. These are the tasks AI is taking over first, and the change already shows up in the staffing mix. Clerical and data-entry jobs fell from 32.3% to 30.6% of agency employment in the year through May 2025, the steepest decline on record for that category. Producer roles rose to 29.1% over the same period.
Taken together, those figures describe agencies moving people into different jobs more than cutting them. The roles that handle paper are shrinking. The roles that face clients are growing. The hiring market leaves little room to do it any other way.
A labor pool that can't carry growth on its own
U.S. Bureau of Labor Statistics; Q1 2026 Insurance Labor Market Study by The Jacobson Group and Aon's Strategy and Technology Group
For agencies that plan growth around new hires, those numbers change the math. A talent pool growing by less than 1% a year can't staff a book that grows faster than that. That makes revenue per employee the more useful planning figure. Here's a rough illustration using the report's two benchmarks (our arithmetic, not a staffing plan): an agency producing $7 million at the traditional rate needs about 40 people. At the AI-native rate, the same revenue would take about seven.
Carriers have already banked their own version of these savings. Over five years, their internal distribution costs fell from about 17.5 cents to 16.5 cents per premium dollar, while commissions paid to brokers held flat. The savings came from carriers' own processing, and broker pay was left alone. For agencies, that means the commission rate isn't where the pressure is coming from. The pressure is on their own cost to serve each policy.
Loss ratios and the bonus check
Contingent and bonus commissions are where book quality turns into cash, and the range is wide. One published carrier schedule in Zywave's report pays 6.74% of premium on a book running a 25% loss ratio and 0.43% on a book at 55% or worse. Quality alone drives a fifteenfold difference.
At the top of the market the sums are large. At the five biggest public brokers, contingent and bonus commissions grew from 5.0% to 6.5% of revenue between 2015 and today. They now account for 15% to 34% of total profit.
Automation connects to this through data. Carriers now review payroll data and site photos throughout the year to spot changes in risk earlier. Brokers whose books hold up under that ongoing review come out ahead. Zywave argues this kind of data discipline builds value before an agency writes a single new account.
When a carrier's bonus can swing fifteenfold on loss ratio alone, the data work that trims an agency's back office can also decide the size of its bonus check. The bonus schedule suggests agency operations leaders should judge that back-office data work by more than the time it saves. They should also look at whether it helps keep books in the better loss-ratio tiers.
Where a licensed human still signs
Zywave also sets a clear limit on automation. AI already reads submissions and compares carrier quotes across the market. Binding coverage is a different matter, and AI can't do it. No state licenses software to sign, and no insurer has yet underwritten the errors of an AI system at scale.
The adoption record backs up that caution. Small-commercial risks have been technically bindable by AI since 2019, yet uptake is still in the single digits. Zywave CEO Martin Simoncic gave carrier and broker executives the same answer: brokers keep their seat at the table, but that is no reason to stand still.
The report also attaches dollar figures to that split. About $63 billion a year in distribution pay goes to judgment-heavy work, compared with $52 billion in the roles most exposed to automation. The larger pool depends on experience: reading the gap between what clients say they need and what they are actually exposed to. That is also the expertise the retirement wave is taking out of the industry.
That complicates the simplest reading of the six-to-one ratio. If the scarce asset is judgment that takes years to build, automation alone won't close the gap. An agency that automates its back office but loses its senior producers could find it has solved the cheaper of its two problems.
Before signing for an AI tool, ask which pay pool it goes after. Tools for document gathering and data re-entry target the $52 billion bucket of automatable work. Anything pitched as replacing binding judgment is selling something no state currently licenses.
Consolidators buy, the middle waits
Zywave sees the market moving along three trajectories. Large national and super-regional brokers are consolidating premium and talent, and they buy capability more often than they build it. Delegated authority premium, meaning the business carriers pass to managing general agents and program administrators, rose over three years from $70.9 billion to $108.7 billion. Retail brokers' ownership share of that wholesale layer roughly doubled over five years, from 15.8% to 30.6%.
Buying capability often looks like a specialist deal. Relation Insurance Services acquired the assets of Wright Insurance Planning, a St. Louis life insurance and long-term care planning firm, effective October 1 on undisclosed terms, Josh Recamara reported in Insurance Business America. The deal brings advanced-markets expertise onto Relation's national platform, meaning life coverage used for estate planning, business succession and executive pay. Advisors with that background are relatively scarce, so acquiring them is faster than training them.
According to Zywave, mid-sized agencies that haven't modernized face the most risk. MarshBerry's 2026 Technology & Governance Report shows a wide gap between trying AI and running a business on it. Nearly three-quarters of brokerage firms say they have adopted AI to some degree, but only 13% use it actively across multiple parts of the business. A survey still counts AI that sits in one corner of a firm, but that kind of use won't lift revenue per head.
For a mid-sized principal, the test is one ratio in next year's budget: revenue divided by headcount, tracked against the $175,000 benchmark. If that number rises while the clerical share of staff falls and the producer share holds, the agency is on the automated side of Zywave's line. If it stays flat while the agency hunts for hires in a market growing by less than 1% a year, the gap will keep widening without anyone deciding it should.
Sources
- A productivity gap is splitting the US brokerage market ↗ · Insurance Business America
- Relation Insurance buys St. Louis life and long-term care ... ↗ · Insurance Business America
Your experts belong here
Every story in MarketScale Business Services starts with a company putting its consultants, practice leads, and account teams on the record. Buyers are already reading this topic. The only question is whose experts they find.
Clients hire the firm whose thinking they have already read, which means fewer cold conversations for your partners.
About the author
The MarketScale Newsroom reports on the companies, technologies, and trends shaping 16 B2B industries. It turns primary sources and expert commentary into clear, useful coverage for the people doing the work.