- Buying medicare advantage leads cost per acquisition is the wrong frame. The number that matters is cost per enrollment net of 90-day chargebacks, and it almost never matches the CPL ranking on your vendor invoices.
- Scope of Appointment (SOA) completion varies sharply by lead source. From auditing MA call-center floors, shared web leads complete SOA far less often than inbound calls, often by 20+ points. That gap alone can flip a $25 lead and a $55 lead on the CPA scoreboard.
- Rapid disenrollment inside the first three months triggers a full commission chargeback from most carriers (per Ritter Insurance Marketing). On a typical MA first-year commission, a 10-point disenrollment gap between vendors can erase the margin entirely.
- Most agencies can’t tie a January chargeback back to the vendor that sold the August lead. The fix is one field: a vendor ID stamped on the enrollment record and carried into the carrier commission file.
- Run a chargeback-by-vendor report every January, April, and July. Your #2 by CPL is usually your #1 by net commission. One “cheap” vendor in the stack is probably net-negative.
Questions this article answers:
- What is the true cost per enrollment on a Medicare Advantage lead?
- What SOA completion rate should I expect by lead type?
- How do I re-attribute a 90-day chargeback back to the original lead vendor?
- When does a $25 shared web lead beat a $55 inbound call?
- What questions should I ask an MA lead vendor before AEP?
- How do DSNP and CSNP lead economics differ from general MA?
Why Buying Medicare Advantage Leads on Cost Per Acquisition Misreads the Vertical
Buying medicare advantage leads on cost per acquisition is the wrong scoreboard for this vertical. Medicare Advantage is the only insurance line where the regulator, the Centers for Medicare & Medicaid Services (CMS), directly shapes lead economics. The Scope of Appointment (SOA), the permission-to-contact rule, and the rapid disenrollment chargeback period all sit between a lead landing in your dialer and a commission staying in your bank account.
The practical effect: the vendor with the lowest CPL on the invoice almost never wins on net commission yield. A cheap shared web lead and a premium inbound call can flip their CPA ranking once you load SOA drop-off and early disenrollment chargebacks into the math.
Most MA floors can’t see this. Their agency management system (AMS) records lead source at first contact, then drops it by the time the chargeback hits months later. The same losing vendor gets re-bought every AEP because the P&L shows enrollments per vendor but never net-of-chargeback commissions per vendor.
This guide gives you three things: a three-column cost-per-enrollment (CPE) model, the single field that lets you re-attribute chargebacks to the right vendor, and the contracting checklist that forces vendors to disclose SOA and persistency numbers before you commit AEP budget. For the broader AEP planning frame, see our Medicare Advantage AEP lead generation playbook and the MA agent CPA benchmarks we publish separately.
The Three-Column CPE Model: Raw CPL, SOA-Adjusted CPL, and Chargeback-Adjusted Cost Per Enrollment
The true cost of an MA lead has three columns, not one. Raw CPL is the sticker. SOA-adjusted CPL loads in the gate the regulator put between you and the sale. Chargeback-adjusted CPE finishes the job by removing the commissions the carrier claws back inside the rapid disenrollment window. Most floors stop at column one.
Here’s the math, in plain form:
- Raw CPL = total lead spend ÷ leads delivered
- SOA-adjusted CPL = Raw CPL ÷ SOA completion rate
- Cost per enrollment (CPE) = SOA-adjusted CPL ÷ SOA-to-enrollment close rate
- Net CPE after chargebacks = CPE ÷ (1 − rapid disenrollment rate)
What is the true cost per enrollment on a Medicare Advantage lead?
The true cost per enrollment is your raw CPL divided by SOA completion, then divided again by close rate, then divided again by one minus the early disenrollment rate. Those three adjustments are what the lead vendor’s invoice doesn’t show, and they’re what determines whether the lead actually paid you back.
What SOA completion rate should I expect by lead type?
SOA completion varies more by lead source than by anything else. From auditing MA call-center floors, the pattern is consistent: inbound calls clear the SOA gate at far higher rates than shared web leads, with exclusive web and live transfers landing in between, and aged leads sitting at the bottom. The actual rates on your floor depend on your text-first workflow, your speed to first dial, and how clean the vendor’s source mix is. The point is to measure them, by vendor, every quarter.
The SOA itself is required under the CMS Medicare Communications and Marketing Guidelines. A Medicare beneficiary has to agree, in advance, to a sales meeting about a specific product type. A lead with no SOA doesn’t enroll. Period.
Typical SOA and disenrollment patterns by lead source
| Lead type | SOA completion | Rapid disenrollment |
|---|---|---|
| Shared web (4–8 buyers) | Lowest of the web sources | Highest |
| Exclusive web | Mid-range | Mid-range |
| Inbound call | Highest | Lowest |
| Live transfer | High | Low to mid |
| Aged (60–90 days) | Lowest overall | High |
Rapid disenrollment is the carrier clawback that hits when a member leaves the plan inside the first three months (per Ritter Insurance Marketing). Carriers typically charge back the full first-year commission. Disenrollment patterns are not flat across the year, and they’re not flat across your vendor stack. Plug your own numbers into the model.
When does a $25 shared web lead beat a $55 inbound call?
Work the side-by-side. Assume a 25% SOA-to-enrollment close rate (a realistic working assumption for an experienced floor) and equal first-year commission per enrollment.
$25 shared web lead, 55% SOA completion, 25% close rate, 10% early disenrollment:
- SOA-adjusted CPL: $45
- CPE before chargebacks: $182
- Net CPE after chargebacks: $202
$55 inbound call, 90% SOA completion, 25% close rate, 3% early disenrollment:
- SOA-adjusted CPL: $61
- CPE before chargebacks: $244
- Net CPE after chargebacks: $252
In that scenario the shared web lead still wins on net CPE, if your floor can move SOA discipline to 55% and your close rate holds. Drop the web SOA to 45% and bump disenrollment to 12% (common for a floor without a fast text-first SOA workflow), and the same lead’s net CPE climbs to roughly $291 and loses the comparison.
This is the math your AMS isn’t doing for you. It’s also the math the cheap vendor is counting on you not to do.
How Do I Re-Attribute a 90-Day Chargeback Back to the Original Lead Vendor?
Re-attribution is the hidden killer in MA lead economics, not the SOA gate. The chargeback hits the carrier commission statement months after the lead first arrived (timing varies by carrier and product, per PIA). By then, most CRMs and AMS platforms have either overwritten or buried the lead source on the original enrollment record. The disenrollment lands as a number in a spreadsheet with no vendor attached.
Why most AMS and CRM stacks drop lead source between enrollment and disenrollment
A typical MA workflow looks like this: a lead arrives with a vendor ID in the URL parameters or the API payload, an agent works it and submits the enrollment through the carrier portal, the AMS pulls the enrollment back in as a new record keyed to the member ID. The vendor ID lives on the lead record. The enrollment record points to the lead by member ID if you’re lucky, and often doesn’t carry vendor source at all.
Months later, the carrier sends a commission file that flags the disenrollment. That file is keyed to member ID and writing agent, not to vendor. The chargeback posts. The vendor that sold you the lead is never named in the data trail that determines whether you re-buy from them next AEP.
The single field that lets you re-attribute chargebacks to the originating vendor
Stamp the vendor ID onto the enrollment record at submission, and again onto the carrier commission file at reconciliation. That’s the whole fix. One field, propagated twice.
In practice:
- When the agent submits an enrollment, your CRM or AMS writes the
lead_vendor_idfrom the originating lead record onto the enrollment record. Not a lookup. A stamp. - When the carrier commission file imports, your reconciliation script joins commission lines to enrollment records on member ID plus writing agent ID, and copies the
lead_vendor_idonto each commission line. - Now every commission credit and every chargeback in your books has a vendor attached. The chargeback report by vendor writes itself.
If you’re on a stack that won’t let you write custom fields onto enrollment records (some legacy MA platforms make this hard), the workaround is a parallel attribution table outside the AMS, indexed on member ID, that holds the vendor mapping. Less elegant. Same result.
How to run a chargeback-by-vendor report in January, April, and July
Run it three times a year: January (post-AEP and lock-in), April (catches OEP movement), July (the pre-AEP commit window). The query is one line once the field is in place: chargebacks ÷ enrollments, grouped by lead_vendor_id, filtered to the trailing 12 months.
What shows up when floors run this for the first time: the #2 or #3 vendor by raw CPL is the #1 by net commission yield. One “cheap” vendor in the stack is net-negative once chargebacks clear. The expensive inbound provider that everyone wanted to cut from the budget is the one carrying the floor’s net margin.
For a related re-attribution problem on the front end, see how server-side conversion tracking for insurance lead buyers solves the same persistence-of-source issue inside the ad platform.
How the 2024 CMS Final Rule and TPMO Disclosure Rules Shifted Compliance Liability Onto the Buyer
Compliance is an economic input in MA, not a separate workstream. The 2024 CMS Final Rule (CMS-4205-F) and the ongoing Medicare Communications and Marketing Guidelines reshaped what lead vendors can do, and shifted real exposure onto the agency buying the lead, not the vendor generating it. When a vendor cuts corners on permission-to-contact capture or Third Party Marketing Organization (TPMO) disclosure, the regulator’s attention lands on whoever’s NPN sits on the enrollment.
What permission-to-contact actually requires under current CMS rules
Permission to contact is what it sounds like. The beneficiary has to affirmatively agree to be contacted by a licensed agent about a specific product line, captured in a form a regulator can audit. The vendor should be able to hand you the proof artifact on demand: the form submission, the timestamp, the IP, the disclosure language the beneficiary saw, and the call recording if it was a phone-sourced lead.
If your vendor can’t produce that artifact for any lead in their stack, that’s the answer. Move on.
How TPMO designation determines who can legally generate your leads
A TPMO is any organization that markets MA plans on behalf of plans or agents, which captures essentially every lead vendor in the market. Registered TPMOs are subject to CMS oversight, recording retention rules, and the standardized TPMO disclaimer that must appear on lead-gen creative.
The practical question for the buyer: is your vendor a registered TPMO, are they running the current disclosure language, and are they retaining call recordings for the window the rules require? If a vendor’s call recording retention is “30 days” or “we delete after billing,” that tells you everything about how seriously they take the regulatory environment they operate in.
Which vendor practices fail a CMS audit
The practices that fail audits, in rough order of frequency: stale TPMO disclaimers (old language on landing pages the vendor forgot to update), permission-to-contact consent bundled with unrelated consents on the same checkbox, generic ad creative that names benefits not actually available on most plans (“free groceries,” “$0 copays”) without the required qualifiers, and co-registration paths where the beneficiary didn’t realize they were opting into Medicare marketing at all.
We’re media buyers and lead-gen operators sharing what we see in the field. This isn’t legal advice. CMS marketing rules are genuinely complicated and change between contract years. Talk to an actual compliance attorney before changing your vendor contracts or consent flows.
The Pre-AEP Vendor Contracting Checklist That Forces Disclosure of SOA and Persistency Data
Send this list to every MA lead vendor in your pipeline before signing any AEP commitment. If a vendor won’t put these numbers in writing, that’s the answer. You’re not asking for proprietary data. You’re asking for the operating metrics any serious MA vendor already tracks for their own floor reporting.
- SOA completion rate by lead type, last 12 months, segmented by AEP, OEP, and lock-in. Annual averages hide the seasonal swing. Same vendor’s SOA completion often shifts meaningfully between October and February as the lead pool changes.
- Rapid disenrollment rate by lead type, last 12 months, segmented by season. Disenrollment patterns are not flat across the year. AEP enrollments disenroll at different rates than SEP enrollments.
- Replacement policy on disenrolled members. What’s the credit, replacement lead, or partial refund?
- Call recording retention window, and the process to retrieve a recording for an audit. “On request within 24 hours” is the floor.
- Permission-to-contact capture mechanism and proof artifact format. Show me the form, the timestamp, the IP, the disclosure language.
- TPMO registration status and current disclaimer language. When was the language last updated against current CMS guidance?
- Exclusive vs. shared definitions and enforcement. A “shared” lead sold to 4 buyers and one sold to 12 are different products at the same price.
- Lead delivery latency SLA. Time from beneficiary submission to lead arriving in your dialer. Anything over 60 seconds on a web lead is a problem.
- Chargeback reporting cadence. Will the vendor share their own chargeback data back to you, broken out by your account?
- Source mix. What channels does the vendor run? Paid search, Facebook, native, co-reg, outbound. Each has a different persistency profile.
- Volume guarantees and ramp curves into AEP. What happens if you over-commit and they under-deliver in the second week of October?
- Termination clause and data portability. If you leave mid-year, do you keep the historical lead-source attribution you’ll need for next year’s vendor decision?
How to Split AEP Budget Across Shared, Exclusive, Inbound, and Aged Leads Once You Know True CPE
Once you have the three-column CPE model running and the re-attribution field in place, AEP budget allocation stops being a copy of last year’s split and starts being a defensible decision per lead type. The framework below assumes you’ve done the math on your own historical data.
When shared web leads beat inbound calls on net CPE
Shared web works when three conditions are true at once: you have spare licensed-agent capacity (so you can absorb the SOA drop-off without leaving leads idle), your floor runs a fast text-first SOA workflow (link sent inside two minutes of lead receipt), and your first-year commission per enrollment is high enough that the lower SOA conversion still clears.
Shared loses when agent capacity is constrained, your floor batches SOA outreach, or you’re selling into a premium product mix where the time-per-sale is too high to justify the lower SOA conversion.
When inbound calls and live transfers win
Inbound dominates when licensed-agent headcount is limited (every agent-minute has to count), you’re focused on premium product or special needs lines, or you’re operating outside the AEP rush when web volume thins out. The inbound call has a high SOA completion rate baked in. You’re paying for the gate to already be cleared.
How do DSNP and CSNP lead economics differ from general MA?
Dual Special Needs Plans (DSNP) and Chronic Special Needs Plans (CSNP) carry a year-round Special Enrollment Period, which fundamentally changes the lead economics. Sales cycles are longer (often 7–21 days from first contact, versus same-day on general MA), eligibility verification is a real step in the funnel, and chargeback profiles differ. DSNP members tend to be stickier than general MA enrollees once enrolled, but the front-end SOA-to-enrollment close rate is lower because eligibility kicks out a meaningful slice of leads.
The practical effect: DSNP and CSNP lead buying tolerates a higher CPL than general MA at the same net CPE, if your floor has the verification workflow to confirm eligibility before burning agent hours.
AEP staffing math: matching lead flow to licensed agent capacity
The constraint that breaks most AEP plans isn’t lead cost. It’s licensed agent capacity. The math:
Max profitable lead intake per day = (licensed agents × productive hours × SOA throughput per hour × close rate) ÷ chargeback-adjusted yield per enrollment
Buy beyond that intake and you’re paying for leads that age out, lose the SOA window, or get worked badly by overstretched agents. Most floors over-buy in the second week of October and pay for it in November with depressed SOA rates across the entire lead pool.
For a related staffing frame, see our Medicare Supplement telesales floor economics piece. The licensed-agent-hour math is the same shape even though the product economics differ.
Frequently Asked Questions
What is the true cost per enrollment on a Medicare Advantage lead?
The true cost per enrollment is your raw CPL adjusted for three downstream factors: SOA completion rate, lead-to-enrollment close rate, and rapid disenrollment rate. A cheap shared web lead with mediocre SOA completion and elevated disenrollment can easily cost more per net enrollment than a premium inbound call with strong SOA completion and low disenrollment. The CPL ranking and the CPE ranking are different numbers.
What SOA completion rate should I expect by lead type?
Inbound calls and live transfers clear the SOA gate at substantially higher rates than shared web leads, with exclusive web sitting in the middle and aged leads at the bottom. The actual rates on your floor depend on your text-first workflow, your speed to first dial, and the vendor’s source mix. Demand season-segmented data from any vendor before you commit, then measure your own numbers quarterly.
How do I re-attribute a 90-day chargeback back to the original lead vendor?
Stamp the vendor ID onto the enrollment record at submission, then propagate it onto the carrier commission file at reconciliation. That single field, carried from lead, to enrollment, to commission line, is what lets you run a chargeback-by-vendor report months later. Without it, the chargeback lands as an anonymous debit and the losing vendor stays in your stack for another AEP.
When does a $25 shared web lead beat a $55 inbound call?
Shared web wins on net CPE when your floor has spare licensed-agent capacity, runs a text-first SOA workflow that closes the gate fast, and keeps rapid disenrollment low. When agent capacity is constrained, SOA is batched, or disenrollment runs hot, the apparent CPL savings on shared web evaporate and the inbound call wins on net commission yield.
What questions should I ask an MA lead vendor before AEP?
Ask for SOA completion rate by lead type segmented by season, rapid disenrollment rate by season, replacement policy on early disenrollments, call recording retention window, permission-to-contact proof artifact format, and TPMO registration status. Any vendor that can’t put those numbers in writing within 72 hours isn’t a vendor. They’re a reseller hoping you won’t ask.
How do DSNP and CSNP lead economics differ from general MA?
DSNP and CSNP carry a year-round Special Enrollment Period, which lengthens the sales cycle and adds eligibility verification as a real funnel step. Front-end close rates are lower because verification kicks out ineligible leads, but enrolled members tend to persist longer than general MA. The net result: DSNP and CSNP can tolerate a higher CPL than general MA at the same net CPE, provided your floor has the verification workflow built.
Build Your 2026 AEP Lead Plan Around Net Commission Yield, Not Sticker CPL
The vendor with the lowest CPL almost never wins on net-of-chargeback commission yield. Most MA agencies can’t see it because their AMS drops lead source between enrollment and the chargeback. Fix the re-attribution field, run the three-column CPE model on your own data, and the right AEP allocation surfaces itself.
If you want help building the CPE model against your actual carrier mix, persistency history, and lead vendor stack, or if you’re shopping for inbound MA call volume that holds up on net commission yield, talk to our pay-per-call team about exclusive MA, DSNP, and CSNP lead routing. We’ll build a custom paid media and lead-buying plan against your AEP volume targets, not a generic vendor pitch.
