“Arborgold already reports revenue by customer.”
It does, and it has no record of the ad that produced the customer, because that happened before the customer existed.
For Arborgold
Export invoices from Arborgold, upload the leads behind them, and rank marketing on completed work rather than on requests for a quote.
CloseRev reads an Arborgold export of invoices — the customer's phone or email, the amount and the date — and matches it against the calls, forms and ads that produced the estimate. Tree care and landscaping quote far ahead of the work, so the gap between lead and revenue is months, and only a long window reads it correctly. Invoices with no traceable lead are reported as Direct / Unknown.
Page updated September 24, 2026.
The gap
Spring fills the book with estimates and autumn fills it with invoices, and no report connects the two back to a channel.
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 record. Phone is the working key in this trade; email coverage is patchier.
What was billed for completed work. Proposals and accepted estimates are not revenue.
When the work was billed, used consistently. The match does not need the estimate date: it credits the earliest lead however long before the invoice it arrived.
Removal, pruning, plant health care, maintenance contract. Their values differ by an order of magnitude.
Step by step
Written for somebody with Arborgold open in the next tab. Report names vary by edition, so each step says what to look for.
One row per invoice with a customer contact detail, the total and a date.
Seasonality is the defining feature here. Any window shorter than a year compares channels across different seasons and calls it a ranking.
Call tracking, the website form, search ads, map listings and any neighborhood or marketplace listing you pay for — combined into one lead file with the source on each row.
A March lead becomes an August invoice. The match credits the earliest lead however far back it is, so the invoice file only needs to reach far enough forward to catch the work.
The join runs on the phone and the email, normalized; repeat work carries the same contact and credits the channel behind the first job.
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 Tree Service page — not from a Arborgold account.
| Channel | Share | Sales | Revenue |
|---|---|---|---|
| Google Ads | 71 | $82,500 | |
| Local service ads | 47 | $55,000 | |
| Meta Ads | 13 | $15,000 | |
| Direct / Unknown | 84 | $97,500 |
Unmatched sales stay in Direct / Unknown. They are never spread across the paid channels to make the total look better.
The argument
A homeowner asks for a quote in early spring. The crew is booked out, the job is scheduled for July, and the invoice lands in August.
A ninety-day attribution window loses that entirely, and every report built on one credits the wrong quarter's marketing or nothing at all.
Reading a full year of invoices against a full year of leads removes the problem rather than tuning around it: the earliest lead is credited however long the gap. The lag itself is not a figure in the report, but the matched-records export carries the lead date and the invoice date on every job.
Channels differ on that lag: an emergency storm call converts in a day, a planned removal in months, and a channel judged on the wrong one looks either urgent or dead.
A large removal can be four figures. A pruning visit is a few hundred. A plant health care program is a small recurring amount several times a year.
Blended, they produce an average that describes nothing, and a channel weighted towards one is ranked as though it produced the others.
Where the export carries a service type, mapped as Product / service, the report cuts revenue by it; the matched-records export pairs type and channel, which is where the channel that is expensive per lead can turn out to be the one producing removals.
Recurring programs deserve the same treatment as any other subscription: the value is the year, not the first visit, and every visit credits the channel that produced the customer, so a year of the program shows in that channel's revenue.
A wind event fills the phone for two weeks. Those calls arrive because a tree is on a roof, not because a campaign worked.
Left in the file undistinguished, a storm inflates whichever channel happened to be running and produces a conclusion that will not repeat.
The honest treatment is to read those weeks separately. Where the export carries a service type or a priority flag, that is straightforward; where it does not, the date range alone usually tells you.
The point is not to discard the revenue — it is to stop it being used as proof that something works when the weather was the cause.
Homeowners with a dangerous limb call. So do the ones who saw the truck, the yard sign or the map result, and none of those leave a form behind.
With no tracked numbers, the lead file covers the web forms only, and web forms are the minority of this trade's inquiries.
Distinct tracking numbers on the website, the vehicles and every paid listing turn a mostly-empty lead file into a complete one, and the cost is trivial next to the budget being judged.
Until that exists the report states the size of the untracked bucket rather than distributing it across the channels that happen to have data.
Fair questions
It does, and it has no record of the ad that produced the customer, because that happened before the customer existed.
The season is short; the year is what makes two seasons comparable. A single season ranks channels against weather.
Then a tracked number on the trucks will tell you how much, and that is worth knowing before you spend anywhere else.
No. It reads a file you exported, so your schedule, crews and customer records are not reachable from here.
Invoices for completed work with a customer contact detail, the total and a date.
An accepted estimate is not money. In this trade the gap between acceptance and invoice is months and some of it never closes.
A year. Seasonality makes anything shorter a comparison of months rather than of channels.
Not as a figure. The matched-records export carries both dates on every job, so the lag is a column away; the report itself credits the earliest lead regardless.
Read separately where a service type or the date range identifies it, so weather is not mistaken for marketing.
Yes, where the export carries a service type. It is usually the most informative split available.
Every visit credits the channel that produced the customer, so the year's program lands on that channel rather than being judged on the first visit.
Repeat work carries the same contact and credits the channel that first produced the customer; a first/repeat column mapped as Product / service shows it apart.
Anything with no traceable lead — mostly untracked phone calls in this trade, and the size of it is the finding.
A contact detail, an amount and a date, plus a service type if you include one. No property addresses required, no notes, no card data. Encrypted in transit and at rest and deleted with the import.
Revenue by source, cut by service type where available, repeat work credited to the channel behind the first job, 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.
Arborgold 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 articleCall tracking answers one question well: which campaign produced the ring. It has no view of what happened after the call ended, which is where the money is decided. This is what the gap costs, and how to close it without replacing the call tracker you already have.
Read the articleA channel judged on its first invoice is judged on the wrong customer. For a business that sees the same households again, the channel that sends loyal customers can look expensive and the one that sends deal hunters can look brilliant. Here is how to compute revenue per customer by acquisition channel at 12 and 24 months from a sales export, read cohorts honestly, switch to margin, and set what you can afford to pay for a customer.
Read the articleStart today
Nothing to install in Arborgold, no API key, and no need to have been tracking anything until now. Last year works as well as this month.