What Actually Predicts an Agent Leaving, By Career Stage

CompsAgent Team · August 21, 2026 · 10 min read

The curve nobody puts on a wall

Most brokerages treat retention as a flat problem: a roster of agents, some percentage of whom will leave this year, addressed with programs aimed at everyone equally. The data doesn't support that shape at all.

Analysis of agent movement published by Relitix and covered by RealTrends — built on Arizona market data — found that switching risk follows a distinct, predictable arc by tenure:

Reading the curve

Year one: the safest agents you have

New agents almost never switch. The research is blunt about why — they're too overwhelmed to consider it. They're learning contracts, chasing a first deal, and have no basis for comparison because they've never worked anywhere else.

This matters because year one is where most brokerages concentrate their retention investment: onboarding programs, mentorship pairings, new-agent training tracks. Those programs are worth running — but they are, in retention terms, spending heavily on the lowest-risk cohort on the roster.

Years two and three: the jump

Then the risk jumps dramatically. Three things converge at once. The agent has closed enough deals to know the job. They've met agents from other brokerages at closings and open houses. And they've now paid enough in splits and fees to have an opinion about what they're getting for it.

This is also the moment they first compare their own toolkit against someone else's — because it's the first time they've seen someone else's.

Years four to eight: the long plateau

Risk declines slowly from the year-three peak but stays above average for another five years. This is the expensive stretch. These agents are producing. They have a client base. They know enough to evaluate a competing offer properly, and they aren't yet anchored by a decade-deep referral network that makes moving costly.

An agent lost in year six takes real pipeline with them — see what a departure actually costs.

After year eight: settled

Past the eighth year, switching risk drops below average and stays there. The referral book is mature, the brand association is established, and the cost of rebuilding both outweighs most offers.

The age curve, separately

Age tracks a similar arc but isn't the same variable — plenty of agents enter the business as a second career at 45.

Age band	Relative switching risk	What's usually going on
Under 30	Rising	Early career, still establishing; mobility is normal at this age in any industry
30–34	High and climbing	Production is building; the split starts to feel like real money
35–42	Peak	Peak earning ambition meets peak financial pressure — this is the danger band
43–49	Declining	Established book, higher switching cost, less appetite for disruption
50+	At or below average	Settled; moves are usually about winding down or a specific opportunity

The overlap is the whole point

Put the two curves together and a specific person emerges: late-30s, somewhere between year three and year seven, producing solidly but not yet a top-of-the-board name.

That's not an edge case. In most brokerages that's a large share of the mid-tier — the agents doing steady volume who aren't in the leadership conversation and don't get the recognition the top producers get. They're profitable, replaceable-on-paper, and statistically the most likely to walk.

Brokerages spend their retention budget on the newest agents, who don't leave, and their attention on the top producers, who mostly don't either. The cohort that actually leaves gets a quarterly email.

What this changes about where you spend

1. Segment your roster by tenure, today. Count how many agents sit in years two through eight. That number is your real exposure. At $15,000–$50,000 per departure, multiply it out and the retention conversation changes tone.
2. Stop treating year one as the retention problem. Onboard well because it builds competence, not because it prevents attrition. It mostly doesn't — those agents weren't leaving.
3. Target the mid-tier deliberately. The elevated-risk cohort is the one least likely to be individually managed. That's a fixable oversight.
4. Match the lever to the trigger. Years two and three are when agents first compare toolkits with peers from other brokerages. That's a competitive-comparison moment, and it's won or lost on things that are visible from the outside — the listing presentation, the comp report, the marketing package. Not on culture, which they can't see from across the street.

The honest caveats

Three things to hold alongside this data:

• It's regional in origin. The tenure and age findings come from analysis of Arizona market data. The shape is likely broadly applicable, but treat the specific turning points as indicative rather than universal.
• Correlation, not cause. Being 38 and six years in doesn't make someone leave. It describes a life stage where a move is both attractive and affordable.
• It won't identify individuals. This tells you which cohort to watch, not which person. Use it for allocation, not for a watch list.

The bottom line

Attrition has a shape, and the shape is knowable. Risk is negligible in year one, spikes in years two and three, runs above average through year eight, and settles after. Age peaks in the late 30s.

If your retention spending doesn't map to that curve — and in most brokerages it maps almost exactly inversely — you're not under-investing in retention. You're investing in the wrong cohort.

Frequently asked questions

At what age are real estate agents most likely to switch brokerages?

Analysis published by Relitix and RealTrends found peak switching risk between roughly 35 and 42 years old. Likelihood rises through an agent's 20s, peaks in the late 30s, and declines back toward average by about age 50.

How many years into their career do agents leave a brokerage?

First-year agents rarely move — they are typically too overwhelmed to consider it. Risk jumps sharply in years two and three and stays above average until roughly the eighth year in the business, after which it falls off.

Which agents should a broker worry about most?

Agents in their late 30s who are between years two and eight of their career. That group sits at the intersection of both risk curves, and it is also where a brokerage's mid-tier production usually lives — experienced enough to be productive, not yet anchored by a mature referral book.

Does new agent onboarding reduce attrition?

Onboarding matters, but the data suggests most retention spend is aimed at year one — the lowest-risk year. The expensive departures happen in years two through eight, when the agent is producing and has enough market knowledge to evaluate alternatives.

Related reading

• The Retention Tool Nobody's Talking About
• Why Agent Attrition Is Costing Your Brokerage More Than You Think
• 7 Reasons Top Agents Leave Brokerages
• Why Every Retention Playbook Skips the Toolbox