How do Medicare agencies prevent chargebacks? Medicare agencies can reduce chargeback risk by improving plan fit, staying in contact after enrollment, monitoring client enrollment status, and acting quickly when a client may be at risk of leaving. Our agency uses a six-touch annual retention schedule combined with book-of-business monitoring tool to make that process repeatable.
By Brinton Atkinson — licensed Medicare agent and founder of G3 Financial. Published 8/22/2026. Last reviewed 08/24/2026.
My Story - Cold. Hard. Truth.
By January the dust was gone, the pendant lights were up, and that island was the nicest thing I owned.
I remember sitting at it one morning in February with a plate of eggs I couldn't touch, doing that thing where you refresh a carrier portal because surely the internet made a mistake. It hadn't. The number at the bottom was negative. Not small. Negative.
And here's the part I want to be honest about: I saw it coming and I didn't act on it.
The first warning wasn't February. It was the end of November, with the Annual Enrollment Period (AEP) nearly in the books. We were writing hard, the boards looked great, everybody was hitting numbers — and I finally pulled our accretion. Members in, members out. Flat. We were sprinting on a treadmill and calling it a marathon. I told myself we'd deal with it after the season, because that's what you tell yourself when there's a week left and the boards are still moving.
Then January came in light. Not catastrophic, just... less than modeled. The kind of light where you squint at it and decide it's a timing issue.
Then February was negative, and the timing issue turned out to be a math issue.
I sat at that beautiful island and ran the same three questions in a loop. How do I pay back the chargebacks. How do I pay for this kitchen. And how do I keep going — because I'd hired ahead of the season, and I had a team who'd bet their year on me after I told them it was going to be a good one.
Then came the part I hadn't modeled at all.
My brokers had their own costs. Call center time, leads, all the stuff you front to have a season. Those bills don't care what happened in January. And when a broker's commissions go negative, they can't pay their bills — so the balance doesn't evaporate. It rolls uphill.
Guess who's uphill.
I was suddenly holding a stack of other people's debt on top of my own, and every one of those balances had a person attached to it. People I liked. People who were also scared and also had families, calling me to ask what we were going to do.
A few months before that, I'd been building projections. Which markets to expand into. What a second season at that volume could look like. And now I was sitting at a quartz island doing an entirely different kind of math — not "what's our growth rate," but "how do I feed my family and pay my bills this month."
Chargebacks don't just take your commission. They take your planning horizon. You go from running a company to surviving a quarter, and it happens fast enough to give you whiplash.
So that's my qualification for writing this. Not that I was naturally good at retention — I obviously wasn't. It's that I got humbled by it while sitting on a very expensive rock, and then went and did something about it. We tore the whole thing down and rebuilt it around what actually holds a book together instead of what just feels productive. Then we built software so it can't sneak up on us again.
And this isn't a war story I've had years to sand the edges off of. It's close enough that I can still tell you what those eggs looked like, which is probably why I'm writing this instead of something more flattering.
What we found is the rest of this article.
Quick note on how this is laid out. Most articles on this topic make you wade through the CMS regulations before they'll tell you anything useful. I'm going to do it backwards. The schedule we actually run is next, because that's the part that saves you money. The mechanics — what causes chargebacks, what they really cost, the recoupment rules, an FAQ — are all further down and worth reading when you've got the time.
But if you've only got four minutes, spend them here.
6-Step Medicare Retention Plan to Prevent Chargebacks
The good news is that preventing Medicare chargebacks isn't hard. The bad news is that it's boring, which is worse, because you can't hire enthusiasm for it.
There's no clever hack here. There's a schedule. Six touches a year, per client, and I'm giving you the whole thing because honestly the industry would be in better shape if more agencies ran it.
Touch 1 — Two weeks before the plan goes live
Text or call. Two weeks out from the effective date. Ninety seconds, tops.
"Hey Gary, it's Brinton. Your plan kicks in Jan 1 — your card should show up in the next week or so. If it doesn't, text me and I'll chase it down. Also: do not throw away the envelope. They hide the good stuff in there."
That's the whole thing. What it actually accomplishes has nothing to do with the card. It re-establishes that you are a person who exists, in the same week his plan becomes real.
Because in about three weeks, a receptionist at a cardiologist's office is going to say seven words to him — "we don't take that plan anymore" — and in that moment Gary either has a recent text from you sitting near the top of his messages, or he has to scroll back to October to find you. One of those outcomes is a phone call to you. The other is a phone call to the 800 number on a mailer.
Touch 2 — Onboarding, right after it goes live
This is the one almost nobody does, and it's the highest-leverage hour you will spend all year.
Walk them through actually using the plan. Set up the gym membership with them — pull up the closest participating location and tell them the cross streets, don't just say "you've got a fitness benefit." Confirm the OTC card arrived, is activated, and that they know what it actually buys. Verify their PCP is showing correctly in the plan's system, not just in the directory PDF. Show them how the dental works. If they're not an app person, hand them the phone number instead and tell them to save it.
Touch 3 — The Open Enrollment Period (OEP) check-in
Roughly half of all chargebacks come down to some version of "this plan doesn't do what I thought it did." Onboarding is where you close the gap between what they think they bought and what they actually bought — while they're still excited, instead of three weeks later when they're already mad and already talking to somebody else.
If you sold them in AEP, reach out again during OEP. January through March, that door swings both directions, and most agents treat it purely as a threat.
We treat it as our window. And we do the thing that sounds insane: we explicitly offer to change their plan.
"You're in the window where we can still make a change if this isn't working — is everything doing what we said it would?"
Yes. You are handing them a loaded weapon and pointing out where the safety is. Do it anyway. Because there are only two outcomes. Either everything is fine, and you just became the single most trustworthy person in his contacts. Or something's wrong — and you get to fix it, instead of the guy from the mailer.
A plan change you write is a retained client and a new commission. A plan change somebody else writes is a chargeback, a lost renewal, and a client who now tells his friends about his agent, who isn't you.
Touch 4 — Quarterly check-ins
Short, human, and anchored to something genuinely useful so it doesn't smell like a sales call from four feet away.
The OTC benefit is the perfect excuse. Most people leave money on the table every single quarter and are legitimately grateful for the nudge.
"Quick reminder — your OTC balance resets at the end of the month. Use it or lose it. Go get the good vitamins, not the store brand."
Then ask how they're doing, and actually listen to the answer. You'll hear about the hip replacement, the move to be closer to the grandkids, the new specialist, the daughter who's "handling his insurance stuff now." Every one of those is a retention risk that you now know about months before it would have shown up on a statement.
Touch 5 — Their birthday
Thirty seconds. Text or call. And the single most important rule of this one: you want nothing.
No plan talk. No "while I've got you." No soft pivot into a referral ask, which they can smell instantly and which turns the nicest touch of the year into the cheapest one.
"Happy birthday, Gary. Hope somebody's making you a cake, and hope it's not sugar-free."
That's it. That's the whole touch.
Here's why it matters more than it looks like it should. Every other contact you have with a client has a reason attached to it — a card is coming, a benefit resets, a window is open. The birthday text is the only one all year where you contacted them for no reason except that they're a person you know. That's the one people bring up back to you long after they've forgotten what their deductible is. That's the one that gets read out loud to a spouse in the kitchen.
It also costs you nothing but a field in your CRM, and you already collected their date of birth on the application. You are sitting on the easiest loyalty touch in the business and most agencies just let it sit there.
Touch 6 — The pre-AEP review, scheduled on the spot
Before AEP, reach out for an annual plan review — and book it on the calendar right then, while you're on the phone. Not "give me a shout in the fall." An actual date and time.
Two things happen. You get in front of your own book before your competitors get in front of it. And the review stops feeling like a renewal he might shop, and starts feeling like a service he already has.
Is that a lot? Yes. Six-plus touches per client per year, across a book of a few thousand people, is real operational weight. That's exactly why most agencies don't do it — and exactly why the ones that do get to keep their carrier contracts.
But here's what you buy with it. People remember you. I've got clients who text me photos from a grandkid's graduation. There's a woman who mails a Christmas card every year with a full page of handwritten updates about her family, and I read every one. Those people do not disenroll. A mailer shows up at their house and it goes straight in the trash, because they already have a guy.
That's the whole strategy. Be the guy.
How We Built Medicare Client Retention Into Our Agency
Running that schedule from memory works until about two hundred clients. Then it doesn't. We know because we tried, and it broke in exactly the way you'd predict — the loudest clients got seven touches and the quiet ones got zero.
The quiet ones are the ones who leave.
So here's what changed at our shop after my February morning:
We made retention somebody's actual job. Not "everyone's responsibility," which is a phrase that means nobody's. One person owns the touch calendar, and it sits on their scorecard the same way production sits on a producer's.
We put the schedule into the CRM as dated tasks tied to each effective date. Not reminders for us to remember — tasks that generate on their own whether anyone's feeling motivated that week or not. Motivation is not an operating system.
We started tracking net accretion monthly instead of annually. Members in, members out, one line, every month. That's the number I should have been watching in September, not stumbling onto at the end of November when the season was already spent. When it goes flat you still have a problem you can fix. When you find out in February, you have a plate of eggs you can't eat.
And we stopped learning about disenrollments from commission statements. This is the big one, and it's the reason I ended up building MAPDTool.
Here's the thing about a chargeback: it's a timing problem before it's a money problem. By the time it hits a statement, that member left 30 to 90 days ago. The conversation you needed to have happened back in January and you weren't invited. At that point you're not doing retention — you're doing an autopsy with a spreadsheet.
MAPDTool monitors your book of business and flags when a client's plan status changes, so you find out while there's still a conversation to have instead of reading about it in an accounting document. You call Gary that week. You find out his cardiologist dropped the plan. And you either fix it inside the same carrier, or you write him into something better yourself — which means he's still your client, and the commission is still yours.
That's the entire thesis, and it's why the schedule and the tool are the same idea: you cannot save a client whose exit you didn't see. The six touches keep you close enough to see most of them coming. The monitoring catches the ones that slip past anyway.
Why Those Touches Work: What Actually Causes Medicare Chargebacks
Now the background — because once you know what actually kills a policy, you'll notice that every one of those six touches is aimed at something specific.
In practice, chargebacks come from a much shorter list than agents expect. It's basically five things, and I'd bet your book on it:
The doctor.
The number one killer, every year, not close. Gary hears from a front desk in January that his plan isn't accepted anymore. Gary panics. Gary does not call you — because you're an insurance guy and he doesn't want to be a bother — so he calls whoever's phone number is in front of him. (Touches 1 and 3 exist for this.)
The drug.
He picks up his prescription and it's $148 instead of $6. He is now furious at a plan he loved eleven days ago. (Touch 2 exists for this.)
The benefit he thought he had.
He believed the dental was $3,000; it was $1,500. He believed the OTC card bought whatever he wanted at Walmart. He believed the flex card covered his glasses and his hearing aids. Somewhere between the commercial with the retired quarterback and your presentation, the number in his head became a different number than the one on the page. (Touch 2 again, and Touch 4.)
The other agent.
Somebody got to him first — at church, at the senior center, at his daughter's kitchen table, through a mailer with a shiny card on it. They promised an OTC benefit but didn't mention all the other points about the plan… This one stings the most, because it's usually not a better plan. It's just a more recent phone call, and your client didn't remember who their agent was. (Touches 4, 5 and 6 exist entirely for this.)
The silence.
The real one. The one nobody puts on a slide. You sold Gary in October, you said "call me anytime" and you absolutely meant it, and then you wrote 340 more applications and he never heard your voice again. By January you aren't his agent. You're a guy who sold him something once. (All six touches exist for this, and the birthday text does more of the work than you'd think.)
Four of those five have nothing to do with the plan you picked. They're relationship problems wearing a product costume.
The rules, briefly. CMS lets carriers recover your commission when a member leaves, and how much depends on when:
Rapid disenrollment — the member drops within the first three months of the effective date. Per 42 CFR § 422.2274, "the entire compensation must be recovered." All of it. Not most of it. After three months — the carrier recovers a pro-rated amount based on the months the member wasn't enrolled.
Know the exceptions, because they'll win you arguments and real money. An enrollment effective October 1, November 1, or December 1 that changes on January 1 is not a rapid disenrollment. Neither is a move out of the service area, a death, gaining Medicaid or LIS, an institutional move, a plan termination, or a move into a 5-star plan.
How Much Medicare Chargebacks Really Cost (It's Not the Number on the Statement)
Most agents look at a chargeback and see one commission. That's the smallest part of it, and it's why the problem stays invisible until it's a crisis.
For 2026, CMS set the national maximum initial MA commission at $694 per member per year, with renewals at $347, and increasing for 2027. (It's higher in CA and NJ at $864/$432, and in CT, PA and DC at $781/$391.) So let's do honest math on one rapid disenrollment in a national state:
Commission clawed back
–$694
Renewal stream you'll never collect
–$1,735 (at $347/year for five years)
Lead cost already spent
–$60 to –$120
Agent comp already paid out
–$150 to –$300
Referral he was going to send you
Unmeasurable, but not zero
Call it $2,600–$2,900 in enterprise value — gone because a front desk said seven words to a guy in January.
Now scale it. Write 1,000 apps in AEP, have 12% rapid-disenroll — which isn't a disaster, that's a normal Tuesday at a lot of telesales shops — and that's 120 members. Roughly $83,000 clawed back immediately, and something north of $300,000 in lifetime value set on fire.
And here's the cash flow trap that got me: the clawback comes due in February and March. Your AEP bills came due in October, November and December. You already bought the leads. You already paid the call center. You already covered the dialer, the licensing, the E&O, the conference booth, the guy who does your Facebook creative, and — if you're a real overachiever — a kitchen. The money is gone. The bill is not.
If you run a team, add one more floor to the basement. Your brokers front their own lead and call center costs. When their commissions go negative, they can't cover those bills — and that balance doesn't disappear, it rolls up to you. So on top of your own clawbacks, you're now carrying debt for people you like, who are calling you - scared for their future, and you're the one who has to decide what happens next.
I've watched agency owners take working capital loans just to survive a chargeback season. I've sat in negotiations over a broker debt balance that started as a rounding error and grew into the number that decided whether the agency lived. It is not a fun room to be in, and everyone in it is polite in a way that makes it worse.
The part that genuinely scares me: accretion. For a long time I treated retention as a margin issue. Nice to have. Something we'd tighten up after the season.
Then I started paying attention to how carriers and uplines actually evaluate an agency, and it isn't just "how many did you write." It's net accretion — members in, minus members out. Your gross production can look beautiful while your net sits at zero. And an agency with flat net accretion is an expensive agency to keep around.
Agencies get cut for this. It's a cold and impersonal letter in the mail with your termination date. The carrier relationship you spent four years building quietly moves to a shop with worse production and better persistency. Nobody ever calls it a termination for low retention. It just stops.
So no — retention isn't a nice-to-have you get to after AEP. It's what determines whether you get another AEP.
Medicare Chargeback Prevention: The Short Version
Strong Medicare client retention is one of the most effective forms of Medicare chargeback prevention. So, run the schedule: pre-effective touch, onboarding, OEP check-in, quarterly nudges, birthday, and a pre-AEP review booked on the spot. Give it an owner and put it in the CRM as dated tasks, because memory doesn't scale past two hundred clients. Watch net accretion monthly, not annually. It's the number that warns you while you can still do something. Know the rules — inside three months the carrier takes all of it; after that it's prorated. Monitor your book so a switch reaches you before it reaches your statement.
None of that is clever, and that's the point. Retention isn't a growth hack. It's being the kind of agent people don't want to leave, done systematically enough that it survives your busiest month.
I still think about that morning, sitting down to eat. But I want to be careful not to leave the story there, because that morning isn't the point of it.
What happened next is that we got obsessive about it. We started testing touches against our own book — which ones actually moved retention and which ones just made us feel busy. Some of the things I was most confident about turned out to be noise. The onboarding hour and the OEP call turned out to be worth more than everything else put together. The birthday text, which I originally thought was a soft, sentimental waste of time, turned out to be one of the stickiest things we do. We threw out what didn't hold and rebuilt the schedule around what did — the six touches above.
Then we built the software, because a schedule only protects the clients you remember and I wanted the book to watch itself. That's MAPDTool. It exists specifically so that no one at our agency ever has to find out about a disenrollment from an accounting document again.
I'm not going to sit here and claim a decade of flawless persistency. This is recent, and I'm still in it. What I can tell you is that the team is still here, the numbers are pointed the right direction, and I'm not doing survival math at my kitchen island anymore — I'm back to doing the other kind. Given where that story started, I'll take it.
You can have the big AEP and keep the business. You just have to decide retention is a real line item before a commission statement decides it for you.
Want the "watch the book" part handled for you? MAPDTool monitors your entire book of business and tells you when a client's plan status changes — while you can still pick up the phone and do something about it. Grab a demo or start a free trial, and bring us your worst February story. We'll trade.
About the author
Brinton Atkinson is a licensed Medicare agent and the founder of G3 Financial, a Medicare telesales agency, and of Spark AI Partners, where he builds automation tools for insurance agencies. He hosts The Insurance Playbook Podcast and built MAPDTool because he saw the writing on the wall. He writes about agency operations, retention, and the parts of this business nobody puts on a slide. Link to Brinton's LinkedIn Profile
FAQ
What is a Medicare chargeback?
A chargeback is when a carrier recovers commission it already paid you because the member disenrolled. Under CMS rules, if the member leaves within the first three months of the effective date ("rapid disenrollment"), the carrier recovers the entire commission. After three months, recovery is prorated by the number of months the member wasn't enrolled.
How much does a Medicare chargeback cost an agent?
The direct hit is the commission — up to $694 for a 2026 national MA initial. The real cost is higher once you add the lost renewal stream ($347/year), the lead cost you already spent, and any agent commission you already paid out. Figure $2,500–$3,000 in total enterprise value per lost member.
What counts as rapid disenrollment?
A plan change within the first three months of the effective date. Key exceptions include enrollments effective October 1, November 1, or December 1 that change on January 1, plus moves out of the service area, death, gaining Medicaid or LIS, institutional moves, plan terminations, and moves into a 5-star plan.
How long is the Medicare chargeback period?
The high-risk window is the first three months after the effective date — that's rapid disenrollment, and the carrier recovers the full commission. But your exposure doesn't end there. For the rest of the plan year, a carrier can still recover a prorated amount for the months the member wasn't enrolled. Practically speaking, treat the entire plan year as the chargeback period, and the first quarter as the part that can actually hurt you.
Can you appeal a Medicare chargeback?
You can dispute one that falls into an exception category, and carriers do reverse those — they're worth fighting. What you can't appeal is a member who left because someone else's phone call was more recent than yours. That one is prevention only.
How do I know when a client is switching plans?
Not from your commission statement — that's 30 to 90 days late. Either you're close enough to your book that clients call you first, or you monitor the book directly with something like MAPDTool. Most agencies need both.
How to Prevent Medicare Chargebacks: A 6-Step Retention Plan