A $27.39 Facebook lead does not buy you a June install. It buys you a form submission, and your crews cannot install a form submission.
I learned to separate those things running accounts where the CPL looked fine and the diary stayed empty. For a seasonal service business, the planning unit is a booked job delivered inside the right window, not a lead bought at the industry median.
Say you run HVAC with two install crews heading into cooling season. You need to know how much advertising will fill their June calendar, when to buy it, and whether your office can turn the resulting enquiries into work. A benchmark answers none of those questions on its own.
I start with the job, then work backward. Here is why that order matters.
A lead benchmark prices the first step, not the finished job
WordStream’s 2026 benchmark puts Meta lead campaigns at a $27.39 median CPL and $1.80 CPC. That is a price for a lead. Not a qualified caller. Not a booked estimate. Not an install your crews can complete in June.
Even as a lead-cost reference, the all-industry figure needs context. Home and home improvement sits at $42.95 per Meta lead, roughly 57% above the $27.39 median. That does not mean every contractor pays $42.95. It means importing the cheaper number into a service-business forecast can make the budget look healthier before anything has happened.
I used to tell clients the benchmark was a starting point. For seasonal workload planning, I now start somewhere else: the calendar. The benchmark can help price the forms later, once I know how many forms the business actually needs.
The click data makes the distinction clearer. Lead-campaign CPC fell 6.25% year over year while CPL fell only 0.98%. Cheaper clicks did not translate into proportionately cheaper forms. Neither metric tells me whether those forms became customers who answered the phone, qualified, and booked.
There are several steps between the advertising result and the business result. Each one can lose people. Price the first step if you want a CPL; measure the whole chain if you want a workload forecast.
Work backward from the delivery date, not the launch date
Take one enquiry from the HVAC example and walk it forward:
- The form arrives Tuesday night.
- Someone calls Wednesday morning and qualifies the enquiry.
- The estimate gets booked for Friday.
- The install lands three weeks out, once parts and crew availability line up.
That is an illustrative sequence, not a promise about HVAC turnaround. Its purpose is to keep the dates separate. The day you buy an enquiry is not the day you deliver the work.
Reported sales cycles vary widely. One home-services benchmark puts plumbing at around a day, pest control at seven days, roofing and electrical near 30, HVAC near 90, and kitchen and bath work up to 180. Your own elapsed time matters more than any of those averages. For budgeting, I need the delay all the way from enquiry to completed job, not just the time it takes to get a signature.

If your jobs take 30 to 90 days to close and complete, buying enquiries inside peak demand can push the resulting work beyond the season you meant to fill. A June advertising report can look busy while June’s crew calendar stays empty.
Advertising prices also move. One benchmark places Q1 CPL near $36.20 and Q4 near $46.48, with holiday competition contributing to the difference. That is not a pricing calendar for an HVAC company. It is another reason not to treat one median as a year-round quote.
I would rather use the business’s actual cycle than borrow a convenient average. Start early enough for the work to reach the delivery calendar, not merely the inbox.
Follow-up determines how many forms become usable demand
Now put the office between the ad and the crew. Someone has to contact the enquiry, establish whether the job fits, and book the next step. Advertising cannot make those steps disappear.
One home-services benchmark reports a 42-minute median response time, with 88% taking longer than five minutes. It also reports higher conversion among two-minute responders. I would not paste those conversion rates into an HVAC forecast. I would ask how quickly this office responds, including after hours, and what happens to the people it misses.
That question matters when comparing channels, too. A contractor guide puts home-improvement Meta leads near $41.26 against $90.92 for Google search, then warns that the cheaper leads need to be judged on cost per booked estimate and cost per job. The lower form price is not the finished comparison.
Here is an explicitly hypothetical forecast:
| Stage | Illustrative outcome |
|---|---|
| Meta spend | $5,000 |
| Forms received | 120 |
| Forms contacted within five minutes | 60 |
| Qualified enquiries | 40 |
| Booked estimates | 20 |
| Jobs won | 8 |
At roughly $41 per form, that produces a $625 cost per booked job. If slow follow-up leaves the same spend producing only four jobs, the cost becomes $1,250 per job. Same CPL. Double the cost of acquiring work.
Those are assumptions, not expected results. Their job is to expose where the forecast depends on the business doing its part. If the office cannot handle the enquiries already arriving, buying more forms is not a repair.
Budget for the follow-up process you can staff, not the one the spreadsheet assumes.
Capacity changes which leads are worth buying
The eight jobs in that example are useful only if the business can deliver them when customers need them. An empty appointment slot and a fully booked week call for different decisions. No all-industry benchmark knows which one you have.
When capacity is tight, the rational move may be fewer, better-fit enquiries even if CPL rises. Meta’s More Volume option optimizes for form submissions. It is doing the job you give it: finding people who submit forms. That is not the same instruction as finding people who answer the phone and accept an estimate.
Higher Intent adds a review step, introducing friction before submission. Qualifying questions about service, zip code, timing, and ownership can also help establish whether an enquiry fits before the office spends time on it. The trade is fewer forms in exchange for a better chance of useful conversations, not a guarantee that every remaining lead books.
For the HVAC company, the questions are concrete:
- Is this a service the crews perform?
- Is the property inside the area they can serve at a margin?
- Does the customer’s timing match the available install window?
- Can the office respond quickly enough to move the enquiry forward?
What the deck calls lead volume, I call names that still need work before they become diary entries.
Capacity decides whether more enquiries are useful. When the diary fills, cutting volume can be better management than celebrating another record week of leads. That is the argument behind Fully Booked? Your Meta Agency Should Cut Leads, Not Chase Them.
Ask for the operating facts before the budget quote
Once the planning unit is a job inside a delivery window, an agency needs more than an ad-spend target. Before I trust a seasonal forecast, I want four sets of facts:
- Conversion: last season’s contact, qualification, booked-estimate, and close rates by service.
- Timing: the delay from enquiry to completed job.
- Service area: the zip codes the business can actually serve at a margin.
- Capacity: the slots the office and crews can staff.
Without those, a budget built on $27.39 is arithmetic without a business inside it. The answers can change whether it makes sense to start eight weeks early or hold spend until capacity opens.
That is the standard I bring to a consultation built around the account: establish what the business needs before treating the platform metric as the objective. Qualification, service area, follow-up, and measurement are also the work behind 8 Jobs Your Meta Ads Agency Must Finish Before You Raise the Peak-Season Budget.
Then comes the plumbing. The form needs to reach the CRM. The office needs a clear response process: instant SMS, then a call within minutes. Booked estimates need to be recorded, and outcomes sent back through Conversions API so the feedback is not limited to form submissions.
A booked estimate still is not revenue. Keep tracking through the won job and its delivery date. Otherwise, you have moved the reporting cutoff without reaching the result you meant to buy.
Management fees do not answer this problem. Typical small-business Meta management is quoted at $300 to $1,000 per month plus ad spend, with percentage-of-spend pricing at higher budgets. Whatever the fee structure, a report that stops at Meta-reported leads leaves the central question unanswered.
I want cost per booked job by service and week, with enough visibility to see whether those jobs fit the available calendar. If the report cannot show that, I cannot use it to plan next season. A polished PDF does not make the missing stages less missing.
Test the model on 25 roofing jobs in May
Now take a new case: a roofing company wants 25 replacements completed in May. The order stays the same. Start with work the crews can deliver, then work backward through the funnel.
Use explicitly hypothetical assumptions:
- Jobs: the target is 25 completed replacements.
- Estimates: at a 30% estimate-to-signed-job close rate, that requires about 83 estimates.
- Qualified enquiries: if 70% attend the estimate, that requires roughly 119 qualified enquiries.
- Forms: if half the raw forms become contacted, qualified enquiries, that requires around 238 forms.
That last assumption combines contact and qualification. It must include both. Contacting half the forms does not mean half the forms qualify, and leaving that distinction out makes the forecast look cheaper than it is.
The figures are approximate planning arithmetic, not a promise that 238 forms produce exactly 25 jobs. I would replace every ratio with the company’s own results before trusting the budget. I would also check that signed jobs can become completed replacements inside May. A signature alone does not fill the delivery target.
Only now does a lead-cost assumption belong in the calculation. At a cost near the $42.95 home-improvement median, those forms cost far more than 238 multiplied by $27.39. The relevant seasonal and audience costs still need to be reflected; the median is a reference, not a quote.
Timing comes next. With a roughly 30-day cycle, buying the forms in the last two weeks of May can push the work into June. Buying earlier gives the enquiries time to move through qualification, estimates, and scheduling. It can also build a retargeting pool: one benchmark reports retargeting CPL at $33.15 against $51.40 for cold audiences. Again, that is context for the plan, not a guaranteed price.
The same spend can therefore produce work outside the month you need it. An acceptable CPL does not rescue the wrong delivery window.
This changes the hiring question. Almost anyone can launch the campaigns. I want the agency that refuses to forecast from $27.39 until it sees the contact process, estimate close rate, and crew calendar. Ask what happens after an after-hours enquiry. Ask what gets sent back through Conversions API. Ask to see booked-job costs by service and week.
Back to those two HVAC install crews heading into June. Their forecast was never 180 forms multiplied by $27.39. It was the 16 installs the calendar could hold, worked backward into estimates, qualified enquiries, and forms, bought early enough to land inside the month and followed up fast enough to survive.
Build next season that way and the benchmark becomes what it always was: an interesting number about someone else’s forms. Your budget buys your workload, or it buys you a spreadsheet of names.

