Repeat Callers in Pay-Per-Call: Detect and Handle Them

By the Tagloud team · November 13, 2026

Every pay-per-call campaign gets the same caller more than once. Some of those callers are real people who got cut off, some are customers checking on a callback, and a few are traffic that should never have been billed. Handling repeat callers in pay-per-call well means agreeing on what counts as a duplicate, picking a window, writing it down as a rule and applying it the same way to every publisher.

This guide is for pay-per-call networks and buyers who route calls from several publishers and want a duplicate policy that is fair, checkable and easy to bill against.

This is a general operations guide, not legal or contract advice. Your insertion orders and written agreements with each buyer and publisher always come first.

What counts as a repeat caller in pay-per-call

A repeat caller is the same phone number calling the same campaign again within a set period. Three details decide whether a call counts as a duplicate:

• Same number. The caller ID matches a previous call. This is the only signal most platforms can check automatically.
• Same campaign or buyer. A caller who rings a Medicare line and later an auto insurance line is usually two separate opportunities. A caller who rings the same offer twice is usually one.
• Inside the window. The second call arrives within the duplicate window you agreed on, such as 24 hours, 7 days or 30 days.

Choosing a duplicate window

The window is how long after a first call a second call from the same number is treated as a repeat. There is no standard length; it depends on how people buy in your vertical.

• Short windows (hours to a day) suit urgent services, where a caller who rings back next week may have a new problem.
• Medium windows (around a week) suit verticals where people shop over a few days, such as many insurance offers.
• Long windows (30 days or more) suit offers with a long decision, such as legal intake, where a second call a few weeks later is usually the same matter.

Pick one window per campaign and make sure your call platform and publishers use the same number.

Why repeat calls happen

Not every repeat is junk. Before you reject anything, sort repeats by cause.

Legitimate repeats

• The call dropped, or the caller hit voicemail or a full queue and tried again.
• The caller was told to call back with a document, a spouse or a policy number.
• The caller is an existing lead or customer checking on a callback.

Repeats that point to a traffic problem

• Many repeats from one publisher, while others send almost none.
• Repeats that arrive in tight bursts, or just past the buffer each time.
• Repeat callers who don't seem to know what they called about.

One pattern alone is not proof of fraud, only a reason to review.

How to write a repeat caller rule

A good duplicate rule is short enough to apply the same way every time. A starting template:

"A call is a duplicate if the same caller number called this campaign within the last 7 days and the earlier call was billed. Duplicates are not paid, unless the earlier call was under the buffer or did not connect to an agent."

Adjust the parts in that sentence to fit your offer:

1. Window: the number of hours or days.
2. Scope: this campaign, this buyer or all campaigns.
3. Which call is paid: usually the first billable call. Some buyers pay the call that qualified instead.
4. Exceptions: dropped calls, calls that never reached an agent, and calls the buyer asked the caller to make.

Keep duplicate handling separate from your quality rules: one asks "have we already paid for this person?", the other "was this person worth paying for?"

Handling billing with publishers

Most duplicate disputes are really disagreements about the window or the scope.

• Put the rule in the agreement. Window, scope, which call is paid and the exceptions, in writing, before traffic starts.
• Show the earlier call. When you reject a duplicate, share the date, time and outcome of the first call, with the caller number masked. "Duplicate of a billed call on Tuesday at 2:14 pm" ends the conversation quickly.
• Handle cross-publisher repeats openly. If a caller first came through Publisher A and later through Publisher B, decide in advance who gets paid. A common choice is to credit the first source, but whatever you choose, apply it to everyone.

For the wider rejection process, see how to dispute junk pay-per-call calls with evidence.

Step by step: a repeat caller process

Step 1: Set the window and scope per campaign

Decide the window in hours or days and whether it applies to one campaign, one buyer or the whole network.

Step 2: Flag every repeat, then review

Mark each call whose number appeared inside the window, rather than dropping it silently. Keeping the record lets you spot patterns.

Step 3: Sort legitimate repeats from problems

Check what was said. "We got cut off" is a callback. A caller with no idea why they dialed is a source question.

Step 4: Count people, not calls

Three calls from one person are one prospect. Build cost per qualified call on unique qualified callers, or repeat-heavy sources will look better than they are.

Step 5: Track repeat share by publisher

Add repeat share next to qualify rate and cost when you grade publishers. A rising share is an early warning.

An example of a duplicate rule at work

Say a network runs one campaign with a 7-day window and three publishers. In a week, Publisher A sends 100 calls with 4 repeats, Publisher B sends 100 with 6, and Publisher C sends 100 with 22. Most of A's and B's repeats mention a dropped call or a callback. Many of C's arrive within minutes of each other. The network pays the first billable call in each case, rejects the rest with the earlier call attached, and opens a source review with C.

These figures are an illustration, not a benchmark.

How Tagloud helps

Tagloud pulls calls from CallGrid, Ringba, Retreaver or any webhook, with an hourly auto-import, and flags repeat callers alongside short calls, hang-ups, wrong numbers and out-of-area callers. Every call also gets a verdict against your own qualification rules, with a reason and quotes from the call, so you can tell a callback from a caller who never meant to call. Caller numbers are encrypted and shown as the last 5 digits only. Publishers are ranked with grades on qualify rate, cost against target and talk time. See how it fits a network on the pay-per-call networks page.

FAQ

What is a good duplicate window for pay-per-call?

There is no single right answer. Match it to how long people in your vertical take to decide: hours to a day for urgent services, about a week for many shopping verticals, and 30 days or more for long decisions such as legal intake.

Should repeat callers ever be paid twice?

Usually not inside the window, but agree on exceptions such as dropped calls, calls that never reached an agent, or callbacks the buyer requested. Whatever you decide, write it down and apply it to every publisher.

Is a high repeat rate always fraud?

No. Dropped calls, full queues and callbacks all create repeats. A share that is far above other publishers on the same campaign, or repeats arriving in bursts, is a reason to review the calls and the source.

Who gets paid when a caller comes through two publishers?

Decide before traffic starts. A common approach is to credit the first billable source within the window. The key is that every publisher knows the rule in advance.