Pay Per Lead vs Pay Per Appointment: The Break-Even Nobody Publishes

You have two quotes. One vendor sells you leads and charges per lead. The other sells you booked meetings and charges per appointment. The second number is several times the first, and there is no honest way to compare them as printed, because they are not measuring the same thing. One is the price of a name and a maybe. The other is the price of a meeting.

The gap between those two prices is the price of the qualification labor — the calling, texting, chasing and scheduling that turns a record into a calendar entry. Under the first model you perform that labor yourself. So the real question is never "which model is cheaper." It is whether buying that labor is worth what it costs you, and that depends on numbers the vendor cannot know and you probably have not measured.

This page gives you one formula, four inputs, and a decision you can make with both quotes open in another tab. Fair warning about everything else on this question: nearly all of it is published by a company that sells one of the two models, which is why none of it shows the arithmetic. Written for marketing agencies, PPC agencies and service businesses selling into owner-operated markets, not enterprise buying committees.

Pay per lead vs pay per appointment: the short answer

Pay per lead bills you when contact information changes hands. You receive a record — a name, a number, sometimes an interest signal — and everything after that is yours to do. Pay per appointment bills you when a meeting lands on your calendar, with the outreach, the replies and the scheduling already done.

Here is the answer, and it is a conditional one: pay per lead is cheaper only if your lead-to-appointment conversion rate clears a threshold set by the ratio between the two quotes. Above that threshold, buying leads produces appointments more cheaply than buying appointments. Below it, every lead you buy costs you more than an appointment would have.

There is no universal version of that threshold, and you should be suspicious of anyone who offers one. It is a property of your market, your follow-up speed and your staffing — not of either model. The rest of this page computes it. One vocabulary note before we start: "pay for performance," "pay for performance lead generation" and "pay per performance" are umbrella terms that cover both models and several others, so the phrase on a vendor's homepage tells you nothing about what you would actually be billed for.

The two quotes aren't in the same unit

Comparing a per-lead price to a per-appointment price is a category error, and both camps quietly benefit from it. The lead vendor looks cheap. The appointment vendor looks premium. Neither has to defend the comparison, because no one is making it properly.

Start from what is actually being sold. In an owner-operated market there is no form fill and no buying committee — the person who answers the phone is the budget, the authority and the need, all in one. Both vendors are working the same finite, largely public set of businesses and phone numbers. They are not sourcing from different machines. They are stopping at two different points on the same line: one hands you the record, the other carries the conversation through to a calendar entry.

Which means the price gap is not the price of a better prospect. It is the price of the work in between. Two variables let you convert one price into the other:

• Your lead-to-appointment rate — of the records you buy, the share that become held meetings. Call it your conversion rate.
• Your qualification labor — the fully loaded cost of producing one appointment from raw records: setter time or your own hours, the dialer and the inbox, and all the calls that went nowhere on the way to the one that didn't.

That second term is the one every published comparison omits, for a simple reason: it never appears on an invoice. It is real money all the same, and leaving it out is what makes pay per lead look free.

(This is the work-stage axis — where the vendor stops. Who builds the target list in the first place is a different question, and a different article.)

The break-even conversion rate, derived

Under pay per lead, your cost per appointment is the price of the leads you had to buy to get one, plus the labor of working them: (lead price ÷ conversion rate) + labor. Under pay per appointment, your cost per appointment is simply the appointment price. Set the two equal and solve for the conversion rate, and the currency divides straight out — which is the useful part, because it means the answer transfers even though prices don't.

Divide everything by the lead price and you are left with two pure ratios:

• k = the appointment price ÷ the lead price. What one appointment costs, expressed in leads.
• w = your labor cost per appointment ÷ the lead price. What your own work costs, in the same units.

The break-even conversion rate is then:

break-even rate = 1 ÷ (k − w)

Beat that rate and buying leads is cheaper per appointment. Miss it and buying appointments is. Both quotes are on the table in front of you, so k takes ten seconds to compute; the whole difficulty of this decision lives in the other two terms.

These are algebraic thresholds computed from 1 ÷ (k − w), not measured conversion rates. No source can tell you your conversion rate — and you can find your own k in ten seconds by dividing the two quotes in front of you.

Read one row out loud. If an appointment costs five times what a lead costs and you count your own time at nothing, you need one lead in five to become a held meeting before buying leads pays. If working those leads costs about one lead's worth of labor per appointment, you need one in four.

Two properties of that table matter more than the table itself.

Your labor dominates when the two prices are close, and barely matters when they are far apart. At a 3× gap, moving from counting your time at zero to counting it at two leads' worth takes the threshold from one in three to every single lead converting — which is to say, pay per lead cannot win at all. At a 20× gap the same change moves the threshold by a fraction of a percentage point. Two buyers with identical quotes can correctly reach opposite conclusions, but only when the quotes are close together.

Counting your labor at zero is almost always a lie. If nobody's actual job is working new records within minutes of receiving them, that labor is not free — it is absent. The leads don't get worked, and your conversion rate collapses regardless of what the table says. Computing what an appointment costs to produce on a given channel is a separate exercise; cost per appointment is a quotient with inputs of its own.

Your close rate can't decide this. Your show rate can.

Close rate cancels. Both models end with a prospect in front of the same seller running the same call. So your close rate divides both sides of the comparison equally and drops out of it. It is a decisive number — it decides whether you should be buying meetings at all, which is the ceiling test, and your maximum affordable cost per held appointment is where that gets computed. But it cannot tell you which model to buy. Anyone arguing for a model on close rate is arguing about a term that appears on both sides of the equation.

That holds on one condition, and the condition is the whole argument: close rate cancels only if the meetings are equally qualified. If a vendor's qualification genuinely changes who is in the room, close rates differ and the term stops cancelling. Which means the premium you are paying is, precisely, a claim about prospect quality — and the written definition of a qualified appointment is the only place that claim can be checked. That is why definitions matter, stated as arithmetic instead of as advice.

Show rate does not cancel. A vendor-booked meeting carries the vendor's show rate; a meeting you booked yourself carries yours. So if the appointment quote bills on booked rather than held, divide that price by the vendor's booked-to-held rate before it enters k, or you are comparing a meeting that might happen against one that did. The distinctions behind that — what booked, held and qualified-held actually mean on an invoice — are worth reading before you sign either contract.

How to find your own lead-to-appointment rate before you sign

The formula is useless without your conversion rate, and nobody tells you how to get it. There are three routes, in descending order of quality.

Measure it retroactively. If you have ever worked a purchased list, an inbound batch, or a prior vendor's leads, the number is already in your CRM. Count how many records produced a held meeting. This is the best estimate you will get, because your actual team produced it at your actual follow-up speed.

Buy a small batch as a measurement, not a campaign. Cap it, make it exclusive, pin the definitions, and size it so one appointment either way is not noise. You are not generating pipeline with this purchase — you are learning a number you would otherwise guess at for the length of a contract.

Or admit you don't have it. A buyer with no conversion estimate is exactly the buyer pay per lead is priced for, because the model transfers an unknown onto the only party who cannot price it. That is not an accusation of bad faith; it is just where the risk sits.

However you get it, an estimate has a shelf life. Categories saturate, and a rate measured a year ago in a pool everyone has since contacted is not the rate you will get now.

The denominator problem

Leads purchased, leads reached, and leads that were reachable at all are three different denominators, and they produce three very different conversion rates from the same campaign. A vendor will quote whichever one flatters the pitch, and so, honestly, will you. Pin the denominator before you trust any number, including your own.

And measure to held meetings, not booked ones. Both models flatter themselves on booked.

What the word "lead" is hiding

Four terms decide what a lead price actually means. Treat each one as a variable in the equation above rather than as a contract detail, because that is what they are — every one of them moves either your effective price or your conversion rate.

Rejection and credit. If you cannot reject an unusable record and have it credited, your real price is the quoted price divided by the share that turn out usable. Suppose a fifth of a batch is unusable — wrong number, out of business, plainly outside your criteria. That is a twenty-five percent price increase that never appears anywhere on the invoice.

Exclusivity. A record sold to several buyers divides your conversion rate, because you are racing competitors to a phone that already isn't answering. "Cheaper per lead" and "sold to three of your competitors" are frequently the same fact wearing different clothes. This conversation is well developed in home-services lead buying and has essentially never been written for B2B agency buyers, so ask the question directly: what does this same record cost exclusive? Put that number in k.

Recency and origin. An aged record converts worse at an identical price — how fast a purchased record decays is its own subject. And a form fill, a scraped list record, and someone who actually replied to an outbound message are three wildly different things wearing one word.

Under pay per lead, you become the party doing the contacting

Here is an asymmetry nothing else on this topic seems to mention. When you buy appointments, the vendor contacts the market. When you buy leads, you do — so the obligations attaching to the party making the calls or sending the texts attach to you, not to whoever sold you the file.

Stated as requirements rather than conclusions: the National Do Not Call rule at 47 C.F.R. § 64.1200(c)(2) is written around residential telephone subscribers, with registry access and scrub timing at § 64.1200(c)(2)(i)(D). The definition of an automatic telephone dialing system at 47 U.S.C. § 227(a)(1) turns on equipment that stores or produces numbers using a random or sequential number generator, which the Supreme Court read narrowly in Facebook, Inc. v. Duguid, 592 U.S. 395 (2021).

Practically, that is three questions before you buy a file: where did this record come from, when was it last scrubbed, is the line type validated. For the operational side, see what the statutes require of the party doing the contacting. General information, not legal advice — talk to counsel about your own situation.

What you keep when the campaign ends

This is the term nobody prices, and the one place pay per lead clearly wins.

Most of a finite market says no, or not now. First contact is one-time-only inventory — you introduce yourself to a given business once, and the record of how that went is worth something. Under pay per lead you keep all of it: every record, every objection, every not-until-Q3. Under pay per appointment you are billed only for the yeses, may never see the contacted-and-declined majority, and at contract end you hold a calendar history rather than a list — while the vendor holds the record of who has already heard from you.

Price that honestly rather than romantically. A residual database is worth nothing without a nurture motion and the discipline to work it, and a decaying list nobody touches is a liability. The point is not that data is automatically valuable — it is that you should find out what you leave with before you start, and that on the way out, the suppression list is the exit asset that matters most.

The "which stage are you at" advice is backwards

The standard advice says buy leads while you are testing a new market, then buy appointments once you are mature. Consider inverting it. Testing is precisely when the conversations carry the learning — which objections recur, what vocabulary the market uses, which sub-segment actually answers. A pay-per-appointment vendor hands you the survivors and keeps the learning. The case where finished appointments genuinely earn their premium is the opposite one: a proven offer, a known market, and spare closing capacity you want filled.

Pay per lead, pay per appointment, pay per sale: one ladder, three stopping points

It helps to see the three models as one line priced at three places rather than as three products. A vendor can stop when a contact record changes hands, when a meeting is booked, or when a deal closes. Each step down the ladder transfers more risk to the vendor — and returns less learning to you, because you see less of what happened on the way.

Pay per sale is the one people ask about and rarely get, and the reason is structural rather than evasive: a vendor pricing per closed sale is really pricing your close rate, and they cannot price what they have not seen. That is why it tends to appear late in a relationship, after a vendor has watched enough of your pipeline to model it, rather than on a first call. If someone offers it cold, ask what they are assuming about your close rate and what happens when the assumption is wrong.

All three get marketed as "pay for performance," which is exactly why that phrase should not reassure you. It describes who carries the risk, not what you get billed for.

Every model gets gamed on the unit it bills

Every comparison of these two models includes a section arguing that the other one has bad incentives. Both are right and both are missing it. The model does not set the incentive; the written definition of the billable unit does. Whatever the invoice counts is what gets optimized. Bill per record and record counts rise. Bill per booked event and booked events rise.

What actually differs is where the resulting quality risk lands. Under pay per lead it lands on a number you can measure yourself: your conversion rate falls, you see it in your own data within weeks, and you re-run the formula. Under pay per appointment it lands inside somebody else's definition of a qualified appointment — which you cannot measure your way out of, and have to read your way out of instead, before you sign. The billing triggers that definition hangs on are covered in detail on the pay per appointment page.

When the honest answer is neither

Two conditions return a third answer that no vendor-published comparison offers, because every one of them assumes you are buying one of the two.

Your ceiling is below both quotes. If the maximum you can afford per held appointment sits under both prices, procurement is not the fix. The levers are deal value, close rate and retention, and no amount of negotiating changes the arithmetic.

Per-unit pricing is the wrong shape for your volume. Per-unit quotes are priced to be profitable at a volume you may not have. Where the volume isn't there, the difference tends to come back as setup fees, minimums and minimum spend — quietly converting a "per unit" quote into a retainer you never negotiated. One tell worth trusting: a buyer who cannot say what happens on the Tuesday after the first meeting books is not ready to price either model.

Run the numbers in this order

1. Compute your ceiling per held appointment. If both quotes sit above it, stop here — neither model is the problem.
2. Normalize both quotes. Divide the appointment price by the vendor's booked-to-held rate if they bill on booked. Confirm what a creditable lead is before you trust the per-lead price, and ask what the same record costs exclusive.
3. Estimate your conversion rate with a named denominator and a sample big enough to survive one bad week.
4. Compute your labor honestly, including your own hours at the rate you would otherwise bill them.
5. Compute 1 ÷ (k − w) and compare it to your estimate.

Then re-run all five at ninety days against real data, rather than treating a pre-signature estimate as settled. Track held meetings in a separate pipeline stage from booked ones, or you will not be able to.

The bottom line

Both vendors work the same market and stop at different points on it. The gap between their prices is the price of the qualification labor in between, and whether buying that labor is worth it comes down to three numbers you own: your conversion rate, your labor cost, and your ceiling. That is the entire decision — and it is why almost nothing published on this question shows the formula, because the formula does not always return the answer the publisher sells.

For the record, TaskBlink books meetings onto your calendar, which puts us on one side of this comparison. It is still the arithmetic we would run with you on a demo call, whichever way it came out: at least 3 booked appointments in your first 30 days or you don't pay, and current pricing lives on our pricing page.

Frequently asked questions