“Our jobs take months, so any report will be out of date.”
The opposite. A long cycle is why period-based ad reporting fails here, and why matching signed jobs back to their original leads is the only honest reading.
For AccuLynx
Export completed jobs from AccuLynx, upload the leads that produced them, and rank marketing on signed work rather than on inspections booked.
CloseRev reads an AccuLynx export of completed jobs — the customer's phone or email, the job value and the completion date — and matches it against the leads and calls that produced them months earlier. Roofing has one of the longest gaps between inquiry and money in this catalog, and that gap is exactly what ad platform reporting cannot see. Jobs with no traceable lead are reported as Direct / Unknown rather than credited to whatever ran most recently.
Page updated September 24, 2026.
The gap
Your ad reporting covers thirty days. Your sales cycle does not. The campaign that produced last quarter's best jobs looks like it produced nothing.
What was sold, to whom, and for how much.
The click, the keyword, the call, and what each one cost.
The file
Three things carry the match: who, how much, and when. Anything else is optional and only changes how the report can be sliced.
On the customer the job belongs to. Roofing leads arrive by phone, by form and by canvasser, and the contact detail is the only thing all three have in common.
The contract value or the invoiced total. Where insurance is involved this is the approved scope rather than the first estimate, and the difference is frequently large.
When the job was signed or finished. Signed date is usually the better one for marketing — it is when the customer decided.
Retail against insurance, or repair against replacement. It is the split that explains why two channels with the same lead cost produce different money.
Step by step
Written for somebody with AccuLynx open in the next tab. Report names vary by edition, so each step says what to look for.
One row per job with a customer contact detail, the value and a date. Look for the export listing jobs with their value rather than a particular report name.
This is the step that matters most in roofing. If your average lead takes ninety days to become a signed job, a file covering ninety days will attribute almost nothing correctly.
Paid search, marketplaces, the door knockers and the referral log, combined into one lead file with each row tagged by its source. A channel missing from it will show up as Direct / Unknown and look like it does not exist.
They behave so differently that averaging them produces a number describing neither. If the export carries the distinction, keep it.
The join is the phone and the email, normalized on both sides. Exact matches count; a phone that agrees only without its country code waits for a person; nothing else is counted.
What comes back
Revenue by channel, the count of sales behind each figure, and an honest bucket for the ones nobody could trace. Sample figures, from the worked example on the Roofing Contractors page — not from a AccuLynx account.
| Channel | Share | Sales | Revenue |
|---|---|---|---|
| Google Ads | 19 | $342,000 | |
| Meta Ads | 9 | $153,000 | |
| Email marketing | 3 | $45,000 | |
| Direct / Unknown | 21 | $360,000 |
Unmatched sales stay in Direct / Unknown. They are never spread across the paid channels to make the total look better.
The argument
A homeowner notices a stain on a ceiling in March, searches, books an inspection, gets three quotes, waits for the insurance adjuster, and signs in June. The money lands in July.
Every ad platform reports on a window shorter than that, so the campaign that produced the inquiry gets credited with an inspection booking at best and nothing at all at worst. Meanwhile whichever campaign happened to run in July looks productive.
Matching a job export that reaches back past the whole cycle is the only way to put the revenue where it belongs. In this trade the correction is not marginal — it can reorder the channels entirely.
A hailstorm produces months of demand in two weeks, and it arrives largely regardless of marketing. A campaign running that week will look extraordinary, and repeating it in a quiet month will disappoint everybody.
This is the clearest case in the catalog for reading a long window and comparing a channel against itself across periods rather than against a single month's leaderboard.
Where the export flags storm or insurance work, mapping that flag as Product / service shows how much of the revenue the weather delivered and how much of it traced to a channel at all. For a ranking without the storm, filter storm jobs out of the file.
Insurance work has a different sales motion, a different close rate, a different margin and a different customer — the approved scope is negotiated with an adjuster rather than sold to a homeowner.
Retail work is sold, priced and closed by your team on its own merits. Marketing that produces good retail leads is not necessarily marketing that produces good insurance claims, and the lead costs differ accordingly.
Averaged together the report says something true of the business and useful to nobody in it. Split, the revenue of each shows on its own, and the matched-records export pairs each job's type with its channel when you need to know which channel to push for which kind of work.
Free inspections are the standard offer, they are easy to count, and they arrive close enough to the click that ad platforms will happily report them as conversions.
The conversion rate from inspection to signed contract varies enormously by channel — a marketplace lead comparing four contractors and a referral from a neighbor do not convert alike — so ranking on inspections ranks on the wrong thing with confidence.
With the inspections as the lead file, the report puts each channel's inspections beside its signed jobs, which is the inspection-to-contract rate — the number that can change the budget rather than confirm it.
Fair questions
The opposite. A long cycle is why period-based ad reporting fails here, and why matching signed jobs back to their original leads is the only honest reading.
Partly, and the report is how you find out which part. Flagging or filtering storm work is the first useful step.
If the marketplace gives you a name and a phone, it is a lead file like any other and it ranks on the same axis as everything else.
No, and it is worth saying why not: a roofing file carries claim detail and photographs, and a tool with standing access to it is a liability nobody needs. It reads a file you exported and holds no credential.
Completed or contracted jobs with a customer contact detail, the value and a date.
Either, consistently. On insurance work the approved scope is usually the more meaningful of the two.
At least twice your average sales cycle. In roofing that is often a year, and a short file is the main reason this exercise disappoints.
Yes, where the export carries the distinction — and you should, because they are two businesses with two different marketing profiles.
Map a storm flag as Product / service to see its share, or filter storm jobs out for a storm-free ranking. Otherwise one storm will make a mediocre channel look like the best decision you ever made.
Yes. Any file with a contact detail and a source is a lead file, and marketplaces rank beside your own channels rather than being taken on faith.
Yes, if the lead file holds the inspections: each channel's inspections then sit beside its signed jobs.
Add the canvassing log to the lead file under its own source name and it ranks like any other channel. Left out, that work lands in Direct / Unknown and your paid channels look worse than they are.
Direct / Unknown, reported as unmatched rather than distributed across the paid channels.
A contact detail, a value and a date. No photos, no measurements, no claim documents, no adjuster correspondence. Encrypted in transit and at rest and deleted with its import.
Revenue by channel from signed work, split by job type where the file allows, and everything unmatched kept visible.
By trade
What the report looks like once the export is in, written for each one.
Other systems
Running more than one system, or comparing? The method is the same and the columns are not.
AccuLynx and the other product names and logos on this page belong to their owners and are shown to identify the software a file comes from. CloseRev is not affiliated with or endorsed by them, and connects to none of them: it reads a file you export.
From the blog
Multi-crew home and field services businesses run on phone calls and job management software, not on web forms. Which means the standard marketing dashboard is measuring the smallest part of the business and missing where the money actually is.
Read the articleIt is easy to compute, easy to improve, and it will happily bankrupt you. A defense of the least popular position in performance marketing: that the metric most teams are held to is actively harmful, and that the alternative is available to anyone willing to do one join.
Read the articleThe channel with the cheapest leads is often the one whose leads almost never buy. Lead conversion rate by source divides the new customers each channel produced by the real, deduplicated leads it sent, counted by the month the lead arrived. Here is the formula, what to take out of the denominator, why small counts lie, how revenue per lead and cost per customer follow from it, and a five-channel worked example.
Read the articleStart today
Nothing to install in AccuLynx, no API key, and no need to have been tracking anything until now. Last year works as well as this month.