Home Services · Marketing & Reporting

Your agency reports leads. A lead that rings out is not a lead.

The monthly report says four hundred and twelve leads at thirty-one dollars each and everyone nods. It does not say how many of those calls were answered, how many were booked, how many were actually your existing customers calling the tracking number, or what the ones that turned into jobs were worth. Cost per booked job by channel is the only marketing number that matters in this business, and almost no shop in the country has it.

So the budget goes to the channel with the best-looking report rather than the best-performing one, spends the same in February as in July against demand that is nothing like the same, and gets its actual results on the twentieth of the following month, assembled by the office manager out of exports.

Four ways the marketing money goes unmeasured

01

Leads are counted, booked jobs are not

Your agency optimizes toward the number it is measured on, which is leads. A form fill that never answers the phone, a call that rings out on a Saturday, and a twelve-thousand-dollar system replacement all count identically in that report. Meanwhile the channel producing the systems might be the one with the worst cost per lead and the best cost per booked job.

02

Review velocity depends on which tech remembered to ask

The two guys who ask get five-star reviews. The other six do not ask. Your Google ranking in the map pack is meaningfully driven by review recency and velocity, so your local visibility is currently a function of which techs happen to have the habit — which is to say, random.

03

You spend the same in February as you do in July

Demand in this trade is violently seasonal and most marketing budgets are flat. So you pay peak competitive rates in peak season, when the phone would have rung anyway, and go quiet in the shoulder months when cost per booked job is at its lowest and your trucks have capacity to fill.

04

Last month's numbers arrive on the twentieth

Revenue by trade, average ticket, membership sales, close rate by tech, marketing spend against booked revenue — pulled out of the field service platform by the office manager and assembled into a spreadsheet, arriving with enough delay that it is history rather than management information.

Joining what you spend to what you actually booked

Attribution is the centerpiece and it is not conceptually hard — it is a join nobody has had time to build. Call tracking tells you which channel produced which call. Your field service platform knows which of those calls became a job and what the job was worth. Connect the two and cost per booked job by channel becomes a real number, on the same denominator, comparable across Local Services Ads, paid search, direct mail, the truck wraps, and the yard signs.

The first month it exists it usually reorders the budget, because the ranking is almost never what the lead-count report suggested.

Review generation gets taken off the techs' memory. Job completion triggers the request, timed for when the customer is most likely to respond, attributed to the tech who did the work so you can see who earns them — and the responses get drafted for a human to post, which is the half of review management everybody skips.

Then budget planning against your own seasonality, using three years of your job data rather than a national average, and a weekly report that arrives Monday morning already explained instead of on the twentieth as raw numbers.

What we build toward

Per channelCost per booked job, not per lead
Every channel measured against booked revenue on one denominator. The reordering this produces in month one is usually the single most valuable output of the engagement.
Every jobReview request, sent same day
Triggered by completion rather than by a tech remembering, attributed to the tech who did the work, with the response drafted for a human to post. Review velocity stops being a matter of personality.
Monday 6amLast week, already explained
Revenue by trade, average ticket, membership sales, close rate by tech, and marketing performance — assembled overnight and narrated, so the week starts with the numbers instead of ending with them.

Attribution needs call tracking configured properly before it means anything, and in most shops it currently is not — wrong numbers on some channels, no dynamic insertion on the website, existing customers polluting the counts. Expect the first few weeks to be plumbing.

What actually gets built

Channel attribution to booked revenue

Joins call tracking to job outcomes in your field service platform so every channel gets a cost per booked job and a revenue figure, with existing-customer calls separated out rather than inflating the counts.

Review request automation with response drafting

Triggered on job completion, timed for response rather than for convenience, attributed to the performing tech, and paired with drafted replies to every review that arrives — the part that actually influences ranking and that nobody has time for.

Seasonal budget modeling

Builds the spend plan from three years of your own job data rather than an industry curve, so shoulder-season budget goes up when cost per booked job is lowest and peak-season spend stops paying premium rates for calls you would have received anyway.

Weekly operating report

Revenue by trade, average ticket, membership sales and attrition, close rate by tech, and marketing performance — assembled overnight and narrated, on Monday rather than on the twentieth.

Google Business Profile maintenance

Services, categories, service areas, photos, and posts kept current across every location, because the profile is what decides whether you appear in the map pack for the searches that produce emergency calls.

Modeled, not claimed

What reallocation does before you spend another dollar

A shop spending a meaningful monthly marketing budget across four or five channels, currently allocated on cost per lead and gut feel.

When cost per booked job gets calculated for the first time, the spread between the best and worst channel is routinely several times over — not twenty percent, several times. One channel is producing jobs at a fraction of what another costs, and the budget has been split roughly evenly because the lead reports looked similar.

Move budget from the worst performer to the best without increasing total spend and booked revenue rises on the same dollars. That is the cheapest growth available to a shop this size, and it requires no new channel, no new agency, and no additional investment — only a number that did not previously exist.

Several x
typical spread between best and worst channel
$0
additional spend required to act on it
Month 1
when the reordering usually happens

A pattern we would expect rather than a result we are claiming. Whether your channels are that far apart is an empirical question, and answering it is the first deliverable.

The three that come up every time

Our agency owns the ad accounts.

Then that is the first thing to fix, and it matters more than anything else on this page. An agency that owns your Google Ads account, your Local Services Ads profile, and your call tracking owns your ability to leave and your history along with it. Get ownership transferred with the agency still in place and still managing — a good one will not object, because they are selling management rather than hostage-taking. If yours resists, you have learned something important for free.

We are not doing anything that looks like fake reviews.

Nor would we build it. This asks every customer after every completed job — no filtering by predicted sentiment, no gating unhappy customers into a private form, which is both against platform policy and the thing that gets shops penalized. The volume comes from asking consistently rather than from asking selectively, and the unhappy reviews that arrive are worth having, because a response to one is read by everybody deciding whether to call you.

I was told tracking numbers hurt your Google ranking.

That concern comes from a real thing, applied too broadly. Publishing inconsistent numbers across directories does confuse your listing, which is why the fix is dynamic number insertion — your real number stays in every citation and in the page source, and the tracking number is swapped in by script for the visitor's session. Your Business Profile keeps its primary number with a tracking number added as a forwarder, which is Google's own supported pattern.

Straight answers

Do we have to change call tracking providers?

Usually not. CallRail, CallTrackingMetrics, and the tracking built into ServiceTitan all expose what is needed. The work is in configuring it properly — dynamic insertion on the site, distinct numbers per channel, existing customers excluded from lead counts — which is where most existing setups fall short.

Can it attribute a replacement that started as a service call two years ago?

Yes, and this is where lead-based reporting misleads worst. A maintenance call in 2024 that becomes a system replacement in 2026 belongs, in some meaningful share, to whatever produced the original call. The model holds a long window, which changes how maintenance plans and low-ticket service marketing look — usually dramatically in their favor.

Does this replace our marketing agency?

No, and a good agency gets better with it. What it replaces is the agency's report as the sole source of truth about performance. Most agencies have never had booked-revenue attribution because it lives in a field service platform they cannot see, and the competent ones are relieved to get it.

How quickly will the weekly report be trustworthy?

The operational half — revenue, tickets, membership, close rates — is trustworthy immediately because it comes straight from your platform. The marketing attribution half takes four to six weeks, because it depends on call tracking being reconfigured and then accumulating enough clean data to be worth reading.

Find out which channel is actually filling your trucks

The Pulse AI Operations Index scores intake, operations, and reporting across a home services business and ranks what to fix first. Marketing is rarely the biggest number — but it is the one where the same dollars can produce more work immediately.