2026 lead-market field note

One test profile. 593 calls.

Published July 19, 2026 · Sources checked July 19, 2026

A new measurement study shows why “just block the number” can fail in a one-to-many lead market. It also gives us a useful discipline: separate what the researchers measured from what a senior-care company’s own policy says.

Scope matters. The study tested health-insurance and Medicare lead-generation sites, not A Place for Mom, Caring.com, or Seniorly. Its measurements explain a broader lead-market mechanism; they are not measurements of those named senior-care services.

The numbers are hard to shrug off.

105health-related lead sites in the main crawl
6,819calls received by the 105 main-crawl profiles
1,053distinct phone numbers used for those calls
593calls received by one individual test profile

Researchers Yash Vekaria, Nurullah Demir, Konrad Kollnig, and Zubair Shafiq instrumented lead forms, monitored controlled phone numbers and inboxes, and bought test leads from three marketplaces. Across the broader experiment they received more than 8,000 calls, 600 texts, and 200 emails. In the main crawl, calls sometimes began within seconds.

Read the April 2026 research paper on arXiv

The mechanism: one form becomes several jobs.

The study found identifying information reaching 73 third parties and observed highly distributed calling. Its opt-out experiments reduced contact, but no method eliminated it completely. Phone-based opt-outs acted fastest; contact could later resume from new numbers or domains.

That architecture is why GeezerKeeper treats a referral cleanup as two layers:

  1. the source layer: revoke contact consent and use the referral service’s current privacy or preference route; and
  2. the recipient layer: contact every provider, advisor, or community you can actually identify.

That split comes from the named services’ own materials. A Place for Mom says contracted senior-care providers may contact a user and provides a preference center. Seniorly says information may be forwarded to participating communities or advisors. Those policies are evidence of a one-to-many workflow; the study’s call totals are not.

See the current human-readable exit routes or use the agent-readable registry.

What to do when the cascade has already started.

1

Name the source.

Find the confirmation email or site you used. Do not guess from an unknown caller ID.

2

Inventory recognizable recipients.

Record the organizations named in calls, texts, voicemails, and emails. Each may need a separate request.

3

Use verified routes.

Prefer official privacy policies, preference centers, and addresses published by the organization itself.

4

Block and report appropriately.

The FTC recommends call blocking and reporting unwanted calls. Caller ID can be spoofed, so keep the date, time, displayed number, and any callback number.

Read the FTC’s unwanted-call guidance

What this evidence does not prove.

Finish the administrative part

Turn the call cascade into a finite checklist.

Build one source request free. If several providers are involved, the $19 browser-only Command Center creates separate tasks, preserves status, exports a ledger, and downloads follow-up reminders. There is no account, subscription, or manual fulfillment queue.

Questions, answered carefully.

Did the 2026 study test A Place for Mom, Caring.com, or Seniorly?

No. It tested 105 U.S. health-insurance and Medicare lead-generation sites. The paper helps explain the general lead-market structure. Named senior-care claims on GeezerKeeper come from each service’s own current policy or terms.

Why can one form lead to calls from several companies?

A referral or lead marketplace can pass an inquiry to multiple providers, advisors, or intermediaries. The source and each recognizable recipient may maintain separate contact systems.

Is blocking one number enough?

It may reduce interruptions from that number, but it does not revoke contact permission everywhere or reach other recipients. The FTC also warns that caller ID can be spoofed.

Primary sources