Retention · 20 July 2026 · 6 min read

Why do insurance agents quit?

Gerard Cavaleri Virtual Closer

Almost every article on this subject quotes agency turnover and new-producer washout in the same breath, as if one number described one problem. They describe two different groups of people leaving for opposite reasons, and merging them produces advice that cancels itself out.

Here is the contradiction, stated plainly. Retention advice for insurance agencies tells you to reduce workload, protect people's time and guard against burnout. It also tells you that new agents fail because they don't make enough calls. Both are true. They are true about different people.

Two populations, two exits

Merging these two into one 'agent turnover' figure is why so much retention advice contradicts itself. Sources measure different people — see the figures below.

Service staff and account managers are salaried, established, and drowning. New contracted producers are commission-only, brand new, and starved. A policy that helps one actively harms the other — protecting a new producer's calendar is precisely the wrong intervention, and pushing activity onto an exhausted account manager is how you lose them.

So separate them before you do anything else. Everything below assumes you have.

Population one: the people who are drowning

The best data on this is the Agent for the Future / Liberty Mutual, 2025 Independent Agents at Work Study, which surveyed more than 1,200 independent agency leaders and team members. The findings are not subtle.

87% say their workload increased over the previous year. Two-thirds are often stressed, over half are mentally and physically exhausted, and 51% report being burned out. Roughly four in ten have considered leaving — and staff reporting burnout are twice as likely to be interested in another role.

Set against that, headline turnover looks almost calm. MarshBerry, Insurance Agency & Brokerage Compensation Report's compensation research put employee turnover in insurance brokerage firms at 16.4% in 2024, easing to 15.1% in 2025 — below the 20–25% typical of financial services generally.

Those two facts are not in tension, and the gap between them is the warning. Roughly 40% of your staff have thought about leaving and about 15% actually did. The rest are still here, and a measured turnover rate is telling you about last year while the burnout numbers are telling you about next.

What actually drives it

The burnout research consistently points at volume rather than difficulty. Hard-market conditions meant more remarketing, more renewals to place, more service work per account — absorbed by a headcount that didn't grow. It is administrative load, not selling, and it is the category most amenable to being taken off a person entirely.

This is where a fragmented system stops being an IT problem and becomes a retention one. Re-keying the same client into three systems, assembling month-end reporting from exports, and hunting for a record across tools is exactly the work that produces the 87% figure. We put numbers on that in what a fragmented tech stack actually costs; the cost shows up in your people before it shows up in your P&L.

Population two: the people who never got started

The other exit looks nothing like this. Roughly 30% of newly contracted agents leave within 90 days — double the annual turnover of the salaried population, compressed into one quarter. Nobody burns out in ninety days. They run out of money.

The sequence is depressingly consistent. An agent is recruited on a realistic-sounding income picture. Contracting and appointment take weeks. Onboarding takes two to three months. During that window they have no book, often no leads, and no method — and they are commission-only, so the income is exactly zero. They quit before the job starts.

Which is why recruiting harder is the wrong lever, and the most expensive mistake in this whole area. If 30% wash out in the first quarter, doubling recruitment doubles the cost of producing the same number of surviving producers. The constraint is not how many arrive. It is how many reach a first sale. We walked through the economics of that in becoming an agency owner.

The three weeks that decide it

Week one: are they in a conversation? Not trained, not certified — talking to a prospect. A new producer who has had ten real conversations by day seven has evidence the job exists. One who has had none has been given a reason to doubt it, and no amount of training replaces that.

Week two: is the lead flow real? "We'll get you leads" is where most recruiting promises quietly fail. If a new producer's pipeline depends on someone remembering to assign them something, it will be thin in exactly the weeks it matters most.

Week three: can they see themselves getting paid? Not a promise — the actual arithmetic on their actual level, on business they can point at. Uncertainty about compensation is corrosive at every level and lethal at zero income.

What each population actually needs

Neither column is a culture initiative. Both are about the shape of the work.

Notice that neither column is about motivation, and neither is a culture initiative. Both are about the shape of the work: one has too much of the wrong kind, the other has too little of the right kind.

The number worth tracking

Most agencies track turnover, which is a lagging indicator of something you can no longer influence. Two better ones, both available today:

Time to first conversation. From contract signed to the new producer's first real conversation with a prospect. Measure it in days. If it is measured in weeks, that is your 90-day washout rate being decided in advance.

Time to first sale. The one number that predicts whether a producer is still here at month six. Segment it by whether they arrived licensed, by lead source, and by who onboarded them — the variation between managers is usually larger than the variation between recruits, and that is a fixable finding.

Neither requires a survey. Both are in your systems already, assuming your systems can answer a question that spans recruiting and production — which, for most agencies, is precisely the thing they cannot do.

Common questions

Virtual Closer works on both halves. For the drowning population it removes administrative volume: one record for conversations, calls, policies and comp instead of re-keying across systems, and background agents that do the chasing nobody owns. For new producers it compresses time to first conversation — automatic lead assignment, AI follow-up that opens leads before the producer picks up the phone, an academy for onboarding you can actually see through, and a rank ladder that shows them the arithmetic on their own level.