Lead fees vs platform fees: the arithmetic
A lead you lose still costs you. Here is how to work out what yours have actually cost, on your own numbers rather than anyone's marketing.
August 6, 2026 · 7 min read
You buy leads. Somebody is now telling you a percentage of the job is a better deal. Before you believe either of us, there is one calculation that settles it, and it is a calculation you can do tonight with your own invoices.
This post is not going to tell you leads are a scam. For some crews they are the cheapest work they can buy. It is going to show you where the line sits, so you know which side of it you are on.
The one number that decides it
A lead model charges you per lead, won or lost. A platform model charges a share of jobs you actually completed and were paid for. Those two prices are not comparable until you convert the first one into the same shape as the second.
The conversion is one division. Cost per acquired job equals cost per lead divided by your close rate.
That division is the whole argument. Close one lead in two and an acquired job costs you twice the lead price. One in four, four times. One in ten, ten times. The price on the invoice never changes; what changes is how many of them you had to buy to get one truck-day of work.
Run it on your own last three months
Three months is long enough to survive one dead week and short enough that you still remember the jobs. Pull the file.
- Total everything the marketplace billed you over the period. Lead charges, subscription, boosts, upgrades, the lot — it all bought the same thing.
- Count the leads that spend actually bought.
- Count the jobs from those leads that you signed and completed. Signed-and-cancelled does not count; you did not get paid for it.
- Divide total spend by jobs won. That is your cost per acquired job.
- Divide that by your average job value over the same period. That is your lead cost expressed as a share of a job — the number to compare.
Most crews have never run step five. It is the only step that makes the two models argue on the same terms, and the answer is frequently not the one either the marketplace or the platform would like.
What the lead invoice does not show
The invoice is the visible cost. The comparison is unfair to your own business if you stop there, because a share-of-job fee has no equivalent of any of the following.
- Sharing. Marketplaces typically sell the same contact to several contractors at once. How many is set by the marketplace, not by you — your plan usually spells it out, and it is worth reading again.
- Contacts you cannot reach. Wrong number, no voicemail, never answers. You still bought it.
- Contacts that were never yours. Out of area, wrong trade, a scope you do not do, or a homeowner who already hired someone last week.
- Refund friction. Most marketplaces have a credit policy, but whether a given dud qualifies is their call, and appealing it costs you an afternoon whichever way it lands.
- Windshield time. The drive to the estimate, the measure, the write-up, the follow-up call. None of it is billable and all of it is spent on leads that never close as well as ones that do.
- Price pressure. When several crews are quoting the same driveway from the same lead, the homeowner is shopping quotes. The ones you win are disproportionately the ones you won by being cheapest, which quietly drags your average job value — the denominator in the calculation above — down.
That last one is the sneaky one. Shared leads do not only lower your close rate. They lower the value of the jobs you do close, and both of those move the arithmetic the same direction.
What a percentage model costs that a lead model does not
Here is the honest other side, and it is not small: on a platform, you do not set the price.
The homeowner saw a number before you ever saw the job, and they agreed to it. That has real consequences.
- You cannot charge more because you are busy this month, or because you happen to be the only crew in that city who does the finish they want.
- You cannot discount to close a job you badly want, either. The lever works in both directions and you do not hold it.
- If the price is wrong for what you find on site, you are raising it through a change process afterwards rather than in the driveway at the kitchen table.
- Your sales ability stops being worth money. If closing is the thing you are best at, a priced-work model takes your best tool out of your hand.
A crew with a genuinely good sales process is giving something up here. Anyone who tells you otherwise is selling. What you get in exchange is on the other side of the trade, and it is worth being precise about what it is.
What the fee looks like on a real job
Below is a representative concrete job run through all four contractor tiers by the same fee engine that pays a crew. The fee is a share of a job that happened.
| Tier | Platform fee | You are paid | Net after costs |
|---|---|---|---|
| Starter | 33.0% | $4,848 | $582 |
| Verified | 30.8% | $5,007 | $701 |
| Pro | 28.5% | $5,177 | $829 |
| Elite | 26.0% | $5,359 | $965 |
That is an instant estimate for a representative job, not a final contract — the numbers on an actual listing are the ones that bind, and every listing shows the payout and the net before you claim it. The rate moves by tier, and the tier moves on completed work and callback rate rather than on what you spend.
The asymmetry that matters most: when you pay
Forget totals for a second and look at timing, because timing is what actually kills small crews.
Under a lead model you pay first and find out later. The charge lands when the lead is delivered. Whether that contact answers the phone, whether the job was real, whether you won it — all of that is settled after the money has already left.
Under a percentage model the fee comes out of a job you completed and were paid for. If nothing closes, nothing is charged. There is no monthly number to cover before you have earned anything.
Say that in cash-flow terms: buying leads converts a variable cost into a fixed one. A slow February still bills you. A rained-out fortnight still bills you. The month your best guy is out with a broken hand, when you have the least capacity to work leads fast, is a month you are still paying full price for them. That shape — fixed spend against variable revenue — is the one that empties a small operating account, and it is a shape a share-of-job fee does not have.
Add the estimating hour to the lead side
The calculation earlier used invoice dollars only, which understates the lead side. Every lead you did not close still cost you a drive, a measure and a write-up.
Do the second pass. Count the estimates you gave that did not become jobs. Multiply by the hours each one took, door to door. Multiply that by what an hour of your time is worth when it is on a job instead. Add it to your lead spend before you divide. That is the real cost per acquired job.
A crew that stops bidding gets those hours back, and they are not small hours — they are weekday hours, in the truck, in traffic. The sibling post to this one, at /blog/priced-work-vs-bidding, costs the estimating hour out properly. Read it before you decide, because it is the part of this comparison that never appears on any invoice from either side.
Where buying leads genuinely wins
The division at the top cuts both ways. A high close rate makes leads cheap, and some crews really do have one.
- You answer inside minutes, every time, including at the weekend. First contact matters enormously on a shared lead and most crews are slow.
- Somebody in your operation actually owns follow-up, and it happens on a schedule rather than when you remember.
- You have the capacity to work every lead you buy, fast. Leads you paid for and never called are pure loss and they are the most common leak.
- Your average job value is high relative to what a lead costs in your trade and market — lead pricing varies sharply by both.
- You are comfortable pricing aggressively to win a shopped quote, and you still make money at that price.
If most of those are true of you, run the arithmetic and it will probably tell you to keep buying leads. Believe your own number over this page. The point of the calculation is that it does not care who wrote it.
What DPOC charges
DPOC charges no lead fees. Nothing to apply, nothing to see the board, nothing to be sent a job.
The platform fee is deducted from a completed job at the tier you hold, which is the arrangement the table above describes. Jobs on the board are already priced and the homeowner has already agreed to the number, so there is nothing to bid and no quote to lose.
One thing to know before it appears on a payout. DPOC holds CSLB #999194. Every job runs under this licence — ours, not a subcontractor's. Crews without their own active licence work under DPOC's and carry a higher platform fee for that coverage, because DPOC's licence and bond are what stand behind the work. Add your own active licence to your account and that premium comes off every job from then on.
The decision, compressed
- Cost per acquired job = total lead spend ÷ jobs actually won from it.
- Express that as a share of your average job value, so it is in the same unit as a platform fee.
- Add the unpaid estimating hours to the lead side, because only one of the two models has them.
- Compare. If leads win on your numbers, keep buying them.
- Then ask the other question the totals do not answer: which of these two bills do you want arriving in a slow month?
Whichever way it lands, you will be running your acquisition cost off measured numbers instead of a feeling, which is worth more than the answer.
Common questions
- Does DPOC charge for leads?
- No. There is no charge to apply, no charge to see the job board, and no charge to be sent a job. DPOC's only contractor charge is the platform fee, and it is deducted from a job you completed and were paid for. A job you did not win costs you nothing.
- How do I calculate my cost per acquired job?
- Take everything a lead marketplace billed you over a period — leads, subscription, upgrades — and divide it by the number of jobs from those leads that you signed and completed. That is your cost per acquired job. Divide that by your average job value over the same period and you have it as a share of a job, which is the only form in which it can be fairly compared to a platform fee.
- Is it cheaper to buy leads or to pay a platform fee?
- It depends entirely on your close rate, which is why nobody can answer it for you. Cost per acquired job is cost per lead divided by close rate, so a crew that closes a high share of leads may well be buying work more cheaply than any percentage model would sell it. A crew closing a small share of shared leads usually is not. Run it on your own last three months rather than trusting either side's marketing, this page included.
- What is a good close rate on shared leads?
- The only close rate worth anything is yours, measured: jobs won divided by leads bought over the same period. Figures quoted in marketplace marketing are not a forecast of your business, and DPOC does not publish an industry number because it has not measured one. What is structurally true is that a shared lead closes lower than a referral, because the same homeowner is talking to several contractors about the same job.
- Why are shared leads so competitive?
- Because the marketplace sells the same contact to more than one contractor — how many depends on the marketplace and the plan. The homeowner is therefore collecting quotes, so speed of first contact and price do most of the deciding. That pushes close rates down and pushes the winning quote toward the cheapest bid, which lowers both the numerator and the denominator in your cost-per-job calculation.
- What do I give up by paying a platform fee instead?
- Price control, and it is a genuine loss. On priced work the homeowner has already been quoted and has already agreed, so you cannot raise the number because you are busy or drop it to close a job you want. If selling is what your business is best at, a priced-work model takes that advantage off the table. What you get back is that you never pay for a job you did not get, and you never spend an unpaid afternoon writing an estimate that goes nowhere.