Cost per signed job is what you paid a source divided by the jobs you actually signed from it. The shorthand is price per lead divided by close rate. A hypothetical $60 lead that closes at 8 percent costs $750 per signed job; the same $60 lead closing at 4 percent costs $1,500. Price per lead is the number printed on the invoice. Close rate is the number that decides whether the source made money or spent it, and shared leads are built in a way that pushes close rate down.
Most owners can quote their price per lead from memory and cannot quote their cost per signed job at all. The second number is the one that runs the year. Below is the calculation, worked through with figures that are made up for illustration, followed by a list of exactly which numbers to pull from your own records so you can replace mine.
Every figure in this table is invented for the example. None of it is a benchmark, and none of it came from a lead vendor’s rate card.
| Line | Hypothetical figure |
|---|---|
| Paid to lead Source A, last quarter | $9,000 |
| Leads received | 150 |
| Price per lead | $60 |
| Leads you reached a live person on | 96 |
| Inspections performed | 38 |
| Jobs signed | 12 |
| Close rate, lead received to job signed | 8% |
| Cost per signed job | $750 |
| Average invoice on those 12 jobs | $3,400 |
| Gross margin after labor, equipment, subs, materials | 45% |
| Gross profit per job | $1,530 |
| Gross profit left after the lead cost | $780 |
At those numbers the source works. It costs $750 to buy a job that produces $1,530 of gross profit, so the lead bill eats 49 percent of the gross profit before a dollar reaches overhead, the truck payment, or you.
Now hold everything constant and cut the close rate in half. Twelve signed jobs become six. Cost per signed job goes from $750 to $1,500. Gross profit per job is still $1,530. The source now leaves $30 per job, and your crews ran dryers and hauled equipment to earn it. Nothing about the price per lead changed. The invoice still said $60. The only thing that moved was how many of those leads turned into signed work, which is why price per lead on its own tells you almost nothing.
Do not use insurance claim averages as your average invoice. The Insurance Information Institute, citing ISO data from Verisk, puts the average homeowners claim for water damage and freezing at $13,954 on a five-year weighted average for 2018 through 2022, excluding Alaska, Texas, and Puerto Rico, at a frequency of 1.61 claims per 100 house-years, roughly one insured home in 60 each year. Fire and lightning averages $83,991 at roughly one home in 425. Those are the whole claims. A mitigation company is one line inside that number, sitting alongside contents, structural repair, and everything else the carrier pays out. Your average invoice is what you invoiced on the jobs that source produced, taken from your own accounting, nothing else.
The frequency figures do explain why this category has so many lead sellers in it. Water losses are frequent and fire losses are severe, so there is real volume and real dollars, and volume plus dollars attracts intermediaries.
Shared leads are sold to several contractors at once. That is the product, not a defect in it. The platform is paid when the lead is delivered, so the incentive is to deliver as many as the buyer list will absorb; you are paid only when the job is signed. Those two incentives point in different directions and the gap between them shows up in your close rate.
Three things happen once a lead is shared:
There is a second detail worth knowing about, which is what it costs to sell leads to you. In the measured US data, “water damage leads” gets about 170 searches a month with a cost per click of roughly $27.35, “water damage lead generation” about 50 searches at roughly $27.50, and “water damage restoration leads” about 90 searches a month. Those are contractors searching, not homeowners. Someone is paying north of $27 for a single click from a restoration owner shopping for leads, and that acquisition cost is inside the price of every lead they later sell you.
Pull these for one source at a time, for one full quarter. Doing it per source is the point; blended numbers hide the source that is losing money.
| Number | Where it comes from |
|---|---|
| Dollars paid to that source | Card statement or vendor invoices, not the dashboard’s spend summary |
| Leads received | Vendor export, then counted again from your own phone log; if the two disagree, use yours |
| Leads you reached a live person on | Call log or CRM |
| Inspections performed | Dispatch schedule |
| Jobs signed | Signed work authorizations, dated inside the quarter |
| Average invoice on those jobs | Accounting, final invoiced amounts |
| Gross margin | Job costing after labor, equipment, subs, and materials, before overhead |
Then compute in this order:
Write down the share of gross profit you are willing to hand over before you look at the answer, not after. Once the number is on the page it becomes negotiable, and the negotiation always ends with keeping the source.
Steps two through four tell you where the loss lives. A low contact rate is a speed and staffing problem you can fix. A high contact rate with a low sign rate is a competition problem, meaning the lead was sold to other people. Those two need different responses, and the blended close rate cannot tell them apart.
Ranking your own site pays nothing in month one. It is slower than buying leads, it produces zero calls while it is being built, and there is no date on the calendar when it starts working. If you switch off purchased leads before your own pages produce calls, you have a gap in the schedule and no way to fill it. Anyone who tells you otherwise is selling you something. The honest version is that you run both for a stretch, and you keep buying leads until your own number gets good enough to stop.
What changes if it works is the shape of the cost. A page that ranks keeps producing calls after it is built, so the cost per signed job from that page falls as the page ages, while purchased leads cost the same on day 400 as on day one. The calls arrive unshared, from someone who dialed you rather than a form that went out to four companies, which is the same close-rate problem from the section above, running in your favor instead of against you.
Where those calls appear is also moving. Whitespark ran 540 manual queries across three US cities and six local service industries and found AI Overviews in 68 percent of local searches, against local packs in 39 percent. Restoration was not one of the six industries studied, so read that as the direction search is heading rather than a restoration figure. The practical version is that pages now have to answer questions plainly enough for a machine to quote them, on top of ranking.
NetPageTwo builds and runs owned search for local service companies. Visibility is $489 a month and covers technical and on-page SEO, AI search optimization, a monthly strategy meeting, and email support. Visibility + Revenue is $1,297 a month and adds conversion work, unlimited landing pages, bi-weekly calls, and Slack. Month to month, 14 days notice, no contract.
Bring last quarter’s numbers from the table above and we will work out your cost per signed job on the call, including the case where the answer is that your purchased leads are fine and you should keep buying them.
Related: restoration SEO work.
