Best Practices May 22, 2026

How Many Times Should Your Agency Call a Medicare Lead? A Data-Backed Cadence

David Castillo
Operations Manager

Most agency principals are losing more lead spend to under-attempting than to over-attempting. The instinct after a few "no answers" is to release the lead. The data says the opposite: 6–8 dial attempts spread over roughly 21 days reaches the contactable population. Stop at three attempts — as a surprising number of agencies still do — and you've left half the contactable beneficiaries on the table while still paying full price for the lead. This article gives the agency operator a defensible, source-by-source cadence framework grounded in the lead-response research that's been around for years but is still routinely ignored on the floor.

The Cadence Math

6–8
Dial attempts that capture roughly 80% of the contactable population on direct-response leads
~5x
Higher contact rate when responding to a lead within 5 minutes vs. 30+ minutes (MIT/Velocify)
21 d
Window over which to spread attempts — concentration causes harassment perception
3
Attempt count where most under-investing agencies stop — leaving meaningful contactable population on the table

Why Under-Attempting Is the More Expensive Mistake

Pull a year's lead invoices and look at the cost-per-lead. The principal who pays $35 a lead and works each one through three attempts is paying $35 to dial three times. The principal who works the same lead through eight attempts is still paying $35 — for a contact rate that's commonly twice as high. The marginal cost of additional attempts is essentially zero (it's dialer time, which you've already paid for). The marginal benefit is contacts that turn into enrollments. Math that ought to be obvious is regularly buried by floor habits and under-built automation.

The well-known Lead Response Management Study (originating in MIT-affiliated research and replicated by InsideSales/Velocify across millions of leads) established two findings that still hold: speed-to-first-call dominates contact rate, and persistence over multiple attempts dominates total contactable population. Most agencies operationalize the first finding (sometimes) and ignore the second.

The Senior Phone Behavior Layer

Medicare leads aren't generic leads. The cadence has to be built on senior phone behavior. AARP technology and connectivity research has tracked steadily increasing mobile-phone usage among Americans 65+ but with continued strong landline usage in the older cohorts. The implications for cadence:

Senior Phone Behavior Patterns That Shape Cadence

1
Lower spam-blocking aggression — older callers are less likely to use third-party blocking apps but increasingly likely to ignore unknown numbers entirely.
2
Voicemail behavior matters — many seniors will return a voicemail from a real-person message but ignore robotic ones. Voicemails count toward attempt counts in TCPA terms.
3
Time-of-day skew — mid-morning (10–11) and early afternoon (1–3) consistently outperform later hours; weekend behavior varies significantly by source.
4
Local-area-code preference — numbers matching the beneficiary's area code answer at higher rates; toll-free numbers underperform.
5
Multi-day persistence — older beneficiaries are more likely than younger consumers to actually pick up if you call back two days later, not the same hour.

A Default Cadence Framework

The following is a defensible default for direct-response Medicare leads. It assumes the lead has provided express consent under TCPA. It assumes you are not on a do-not-call list. As we covered in our TCPA compliance guide, every cadence has to be evaluated against the consent record for each lead.

Default 7-Touch, 21-Day Cadence for Direct-Response Medicare Leads

Touch Timing Channel / Window
1 Within 5 minutes of lead arrival Live dial; voicemail if no answer
2 Same day, mid-afternoon Live dial
3 Day 2, mid-morning Live dial; voicemail with first-name reference
4 Day 4, early afternoon Live dial
5 Day 7, mid-morning Live dial + SMS if consented
6 Day 12, early afternoon Live dial
7 Day 21, mid-morning — final attempt Live dial; clear voicemail; route to nurture

Tuning the Cadence by Source

The default is a starting point. Operators tune it source by source based on the actual contact-rate curve from the dialer reports. A T65 internet lead from a high-intent vendor will hit on touch 1 or 2 most of the time and benefit very little from touch 6 or 7. A direct-mail-driven inbound that left a callback request behaves differently. A purchased aged-lead list will look like a flatter curve where touches 4–7 deliver disproportionate contact.

How to find your per-source contact-rate curve

For each lead source, plot cumulative contact rate against attempt number across at least 90 days of dial data. The curve will plateau at a point unique to that source. Set the cap at the inflection point plus one attempt. Sources where the curve is flat by attempt 4 don't need an 8-touch cadence; sources still climbing at attempt 6 may justify one or two more.

The Compliance Ceiling on Cadence

Even when you have express consent, the FCC's TCPA framework treats persistent, harassing patterns of calls as actionable. Most state telemarketing statutes layer caps on attempts per day or per week. Federal Do-Not-Call rules and consent revocation override any cadence the moment the lead asks to be removed. Practically, the cadence has to be enforced as a hard cap by automation, not as a guideline producers follow when they remember.

That means the dialer logic, not the agent, decides when the next attempt fires. Once the cap is reached, attempts stop — even if the producer wants "one more try." Once consent is revoked or the lead asks to be removed, every subsequent attempt is suppressed across all queues, not just the one the producer was working.

Touch Counts vs. Touch Attempts: A Distinction Operators Get Wrong

A single agent picking up the lead and dialing twice in three minutes is not two cadence touches; it's one. Cadence is logic about when to attempt next, not how many times the dialer engaged the line. When auditing your floor's cadence adherence, look at "touches with at least N hours between them" rather than total dial events. Otherwise, an aggressive producer can burn through the cadence in 20 minutes and trigger TCPA exposure for the agency.

SMS, Voicemail, and Multi-Channel Cadence

When the lead has consented to SMS, layering text messages between dial attempts increases contact rate without significantly increasing perceived harassment. The default integration: SMS on touch 5 (day 7) reminding the lead about the inquiry; SMS on touch 7 with a "we'll close out the request unless we hear back" framing. Voicemails on touches 1, 3, and 7 with progressively more specific reasons to call back generally outperform leaving voicemail every time.

Cadence Auditing for Agency Principals

Weekly Cadence Audit Checklist

  • Per-source contact-rate-by-attempt curves — reviewed weekly; cadence cap adjusted if curve has shifted.
  • Speed-to-first-call distribution — should peak under 5 minutes for inbound leads; tail beyond 30 minutes flagged for ops.
  • Cadence-cap respect rate — percentage of leads where cap was respected; manual overrides logged and reviewed.
  • Consent-revoke propagation time — measured in seconds; revocations should suppress dialing instantly across all queues.
  • Time-of-day adherence — calls outside the 8 AM–9 PM window blocked; analyzed by region for state-specific overlays.

Connecting Cadence to Cost Per Enrollment

The bottom-line metric is cost per enrollment. A lead that closes after seven attempts costs the same to acquire as one that closes after one. Adding the right cadence can move CPE meaningfully without increasing lead spend. Operators who track cost per enrollment per source, compared against contact-rate curves, find that the highest-CPE sources are usually the ones being under-attempted, not the ones being over-attempted.

Key Takeaways for Agency Operators

  • Under-attempting is more expensive than over-attempting — the marginal cost of additional attempts is dialer time you're already paying for.
  • Default to 6–8 touches over 21 days — tune up or down based on per-source contact-rate curves.
  • Speed-to-first-call dominates contact rate — aim for under 5 minutes on inbound leads.
  • Senior phone behavior is real — mid-morning and early afternoon, local-area-code, multi-day persistence.
  • Enforce caps in automation, not in agent training — producers will exceed caps unless the system prevents it.
  • Audit weekly — cadence is a metric, not a setting; sources change, regulations change, behavior changes.

The agencies with the lowest cost per enrollment in Medicare aren't the ones with the cheapest leads. They're the ones whose cadence is built into their dialing automation as a hard constraint, tuned per source against actual contact-rate data, and audited weekly. That posture turns lead spend from a guessing game into a managed input — and frees the principal to make decisions about scale instead of decisions about what to do with the leads already in the system.

Cadence That Enforces Itself

AgentTech Dialer's automation builder lets agency principals configure the cadence and lets the system enforce it — stopping attempts at the cap, respecting consent revocation across queues, and keeping DNC and TCPA exposure bounded.

Try AgentTech Dialer Now

References & Authoritative Sources

The information on this page is supported by the following official and authoritative sources.

  1. 1
  2. 2
  3. 3
  4. 4

Related Articles

May 21, 2026

Mid-Year Medicare Strategy

April through September is when smart agencies recruit, train, and optimize tech. A month-by-month tactical playbook for agency owners.

May 17, 2026

AEP Mock-Call Program

Agencies running 20+ hours of pre-AEP mock calls see 15–20% higher AEP close rates. A program design framework for agency principals.

May 13, 2026

Medicare Retention Playbook

Industry persistency averages roughly 78%; top agencies hit 90%+ via structured renewal calls, plan reviews, and ANOC outreach. The full playbook.

Last updated: