What Is Pay Per Call Marketing? A Practical Guide for 2026

Pay per call is a performance marketing model where an advertiser pays for a qualified inbound phone call rather than a click, an impression, or a form fill. The unit being bought is a conversation with a person who picked up the phone on purpose.

That distinction sounds small. In practice it changes almost everything about how campaigns are measured and priced.

Why the phone call is a different unit

A click is a signal of curiosity. A form fill is a signal of mild interest, often supplied by someone who filled in four other forms in the same session. A phone call is someone who has decided the fastest route to what they want is to talk to a human right now.

That intent gap shows up in close rates. Across the verticals we operate in, inbound calls close at several times the rate of comparable web leads. The caller has self-selected for urgency.

It also shows up in how quickly the outcome is known. A form lead might sit in a queue for two days before anyone learns whether it was any good. A call resolves in minutes.

What makes a call billable

This is the part that determines whether a pay-per-call programme works or turns into a monthly argument.

In a well-run network, the advertiser defines billability explicitly, and that definition is visible to publishers before they send any traffic. A typical definition includes:

• Minimum connected duration. Usually 60 to 120 seconds. Long enough to establish the caller is real and relevant, short enough that a genuine caller who hangs up on hold is not punished.
• Geography. Down to state, ZIP, or area code, matched against where the buyer is licensed or has service coverage.
• Qualification criteria. Screened in the IVR before transfer — homeownership for solar, debt threshold for debt relief, diagnosis and exposure window for mass tort.
• Duplicate suppression. A configurable lookback window so the same caller does not generate two billable events.

If any of those rules are ambiguous, you will spend your time on disputes rather than optimisation. Publish them, enforce them at the routing layer, and reconcile against carrier call detail records rather than a self-reported log.

The economics, honestly

Pay per call carries a higher unit price than click or form-fill buying. A call in insurance might cost $18 to $95; in legal it can run past $400. Compared with a $4 click, that looks expensive.

The relevant comparison is not per unit — it is per outcome.

Consider an agency buying clicks at $4 with a 3% call rate and a 20% close rate on those calls. That is roughly $667 in media per written policy. The same agency buying calls at $46 with a 74% billable rate and a 27% close on billable calls lands closer to $230.

The numbers vary enormously by vertical and by how well qualification is set up. The structural point holds: moving payment closer to the outcome shifts risk toward the party who can control it. Publishers who can generate qualified callers get paid more; publishers who cannot, do not.

Where it goes wrong

Three failure modes account for most disappointing programmes.

Duration gates set too high. A 180-second minimum on a vertical where genuine calls average 90 seconds means you are not filtering for quality, you are filtering for stamina. Publishers respond by padding calls, and the calls get worse.

No qualification before transfer. If your agents are spending the first four minutes establishing whether the caller is eligible, you are paying twice — once in media, once in agent time. Move that screening into the IVR.

Capacity ignored. Buying volume your team cannot answer means paying for calls that ring out. Concurrency caps are not a limitation, they are a cost control.

Is it right for you?

Pay per call fits when the sale genuinely happens on the phone, when your team can answer during the hours you buy, and when you can articulate what a good caller looks like specifically enough to encode it in rules.

It fits badly when the purchase is self-serve, when call volume would swamp a small team, or when nobody can say what disqualifies a caller. In that last case the problem is not the channel — it is that the qualification criteria do not exist yet, and no amount of media will substitute for defining them.