Pay Per Lead vs Commission: Which Costs You Less

A pay per lead platform charges you when a lead is delivered, win or lose. A commission platform charges a share of work you actually completed. Everything else about the two models is detail, and that single difference is why two providers can look at the same platform and honestly disagree about whether it is expensive.

The disagreement is rarely about arithmetic. It is about which month each of them is describing. Ask a provider in a busy month and the commission is the thing they resent. Ask the same provider in a slow one and the lead invoice is the thing that hurts. Both are telling the truth about their own year.

So the useful answer to "which costs less" is not a number. It is a rule, and the rule turns on two things you already know about your business: how many jobs you finish in a month, and how much you charge for them.

Naibrly's fee table at marketplace fees groups the platforms it tracks by which model each one runs, with a checked date on every figure and every competitor number labeled as an estimate. This post is the reasoning behind that table. The numbers stay there rather than being retyped here, where they would quietly go stale.

What a lead fee charges you for

You pay when a lead is delivered, win or lose. That is the whole mechanism, and it is worth sitting with, because it means the amount on the invoice is not the price of a job. It is the price of a chance to bid on one.

Three consequences follow, and providers tend to meet them in this order.

Your cost per won job is a multiple of the lead price. If you win one in four of the opportunities you pay for, four charges sit behind every completed job. That multiplication is the most common way a lead budget gets underestimated, and it has its own post: a lead price is not a cost until you divide it.

The cost arrives before the revenue, and sometimes instead of it. A per-lead charge is due on delivery. The job it might become is weeks out, if it happens. In a month where quoting is busy and closing is not, the two sides of that ledger do not line up.

Your close rate is the lever, and nobody else can pull it. Because the cost is lead price divided by close rate, moving from one in five to one in four cuts your cost per won job by a fifth without a single price changing anywhere. That is real, and it is the strongest argument for the model: a per-lead cost rewards a provider who answers fast and quotes well. It is also the reason the model punishes a provider who does not, which is the same fact viewed from the other side.

What a commission charges you for

A percentage or spread comes out of completed work. Nothing is charged for a job you quoted and lost.

That removes the close-rate multiplier from the cost side entirely, which is the change most providers feel first. It does not remove the close rate from your business, and it is worth being clear about that. Quoting work you do not win still costs you the hour it took to quote. What changes is that the hour is the whole cost, rather than the hour plus an invoice.

Two other consequences matter as much and get discussed less.

The bill tracks your revenue. A slow month produces a small fee because it produced little completed work. For a seasonal trade, that is not a rounding difference. It is the difference between a quiet February that is merely quiet and one that is expensive.

A big job costs more in absolute dollars. A percentage of a large ticket is a larger number than a percentage of a small one. On a flat per-lead charge, the large job costs the same as the small one.

The number that decides it is your monthly volume

A commission does not have to come with a monthly plan fee. Naibrly's does, and that pairing is where a real break-even point comes from.

A plan with a monthly fee plus a rate has two parts. The variable part behaves the same at any volume: a percentage of completed work is a percentage of completed work. The fixed part behaves very differently. Divided across two jobs it is heavy. Divided across thirty it is close to noise. Every job you add makes the fixed part cheaper per job, permanently, with nothing to negotiate.

A per-lead cost has no fixed part to spread. Adding jobs adds charges in proportion. The cost per won job at thirty jobs a month is the same cost per won job you had at three.

Two costs that move differently as volume rises have to cross somewhere, and that crossing point is a specific number of finished jobs a month. It is published per platform on the marketplace fees page, refreshed with the underlying fee research rather than typed once and left.

Two things about that table are worth flagging before you read it. Some rows have no break-even volume at all, because no amount of volume changes which model is cheaper for that platform. At least one has none because the competitor's cost behaves as a floor rather than a curve. Those rows are on the page instead of being quietly dropped, which is the only condition under which a break-even table is worth anything.

Job size cuts the other way

Volume favors the model with a fixed component. Job size favors the model without a percentage.

A commission takes more from a larger ticket by design. A flat per-lead charge does not. So above some job value, an estimated lead cost becomes the cheaper of the two, and the fee calculator states roughly where that crossing point sits under a fixed close rate, along with the caveat that matters more than the figure: lead prices climb with job size. A page that quotes one crossing point and stops has told you half of it.

The practical version is that the two models suit different shapes of business. High volume and modest tickets lean one way. Occasional large installs lean the other. A trade that does both is genuinely a harder call, and anyone who tells you otherwise is selling.

What each model does to a slow month

This is the part providers ask about after they have already run the arithmetic, and it is the part the arithmetic does not capture.

Under a per-lead cost, a slow month is a month where you spent on opportunities and finished less work. The spend is not recoverable and it is not proportional to what you earned. Under a commission, a slow month is cheap, because the fee only exists where completed work exists. If there is a monthly plan fee, that is the floor, and it is a floor you can see in advance.

Whether that matters depends on your reserves and your season. It is not a cost difference so much as a variance difference, and variance is what actually ends small trades businesses.

What each model does to your pricing

Acquisition cost gets priced into quotes, whether or not you do it deliberately.

If winning a job reliably costs you a known amount before you start, that amount is in your number. It has to be. Providers who have never divided their lead spend by their won jobs are usually carrying it anyway, in a margin that feels thinner than it should for reasons they cannot name.

This is the quiet argument for knowing your cost per won job under either model. It is not so you can be angry about a platform fee. It is so the figure in your quote is deliberate.

Run the comparison on your own numbers

Four steps, and none of them need a spreadsheet.

1. Take one platform and one full quarter. Add up everything it charged you, including any plan or membership fee.
2. Count the jobs you actually completed through it in that quarter. Not opportunities, not quotes. Completed and paid.
3. Divide. That is your real cost per won job on that platform, and it is almost always higher than the number you would have guessed.
4. Compare it against the per-job figures on the fee calculator, which lets you pick a platform and set your monthly volume.

If you want the same comparison already assembled for a specific platform, the per-platform pages carry it with the estimate hedges and the checked dates attached: Naibrly versus Thumbtack and Naibrly versus Angi.

Where Naibrly sits

Naibrly runs a commission model with a monthly plan, and the reason is the slow month rather than the busy one. Nothing is charged for a lead, a quote, or a job that does not happen. Core is 5% of completed jobs plus $15 a month; Starter is 10% of completed jobs plus $5 a month. Those are published rates, not estimates. On a $100 job at 20 jobs a month, Core comes to 5.75% of completed revenue once the plan fee is spread across the month.

What that buys you is a cost you can predict from work you have already been paid for, which is a different thing from a cheaper cost, and worth separating. At low volume a per-lead platform may well cost you less. The marketplace fees page publishes the volume at which that stops being true for each platform, including the ones where it never does.

If the arithmetic works for the way your month actually looks, start as a provider.