First-party data · 12 accounts · 90 days
Home services Google Ads: the real cost per lead.
Across 12 home-services accounts I manage, cost per lead over the last 90 days ran from $72 on a septic account to $258 on a roofing account, with a median of $138. The spread is real: a $300 drain clear and a $15,000 roof can carry the same ad, but not the same cost per lead. The harder finding is this. Four of the 12 accounts had conversion tracking that was inflated or entirely missing. In a third of the book, the cost per lead on the dashboard was fiction.
Last updated June 2026 · Google Ads platform data, anonymized and aggregated
- 12 Home-services accounts
- $70,142 Managed spend, 90 days
- $138 Median cost per lead
- 1 in 3 Had broken conversion tracking
01 The method
My book,
not a survey.
This is not an industry survey or a number I found in someone else's report. It is the real Google Ads data from twelve home-services accounts I personally manage: HVAC, plumbing, electrical, roofing, fencing, septic, landscaping, lawn and tree care, pest control, painting, exteriors, and paving. The window is the last 90 days. Every figure traces to the account's own reporting.
I anonymized and aggregated it. No business is named, no account is identifiable, and I excluded the accounts with broken tracking from the cost-per-lead numbers so the benchmarks are not poisoned by bad data. The broken accounts get their own finding below, because the breakage is the most useful thing in the dataset.
02 The finding that matters
A third of the accounts were
counting the wrong thing.
Before I trust a single cost-per-lead number, I check whether the conversion is real. In this book of twelve, it often was not.
-
3
accounts counted every phone call as a conversion. Their reported conversion rates were 82%, 83%, and 106%. One logged more conversions than it had clicks.
-
1
account had no conversion tracking at all. Zero conversions recorded on $4,166 of spend. The dashboard could not tell you what a lead cost because it was not counting one.
-
1
more was likely undertracking: a 2.4% conversion rate on $10,000 of spend, low enough that calls were almost certainly going uncounted.
So four of twelve were clearly broken, and a fifth was suspect. By a generous count, nearly half the book had conversion data I would not bid against. These are not bad businesses. They are normal accounts, set up the normal way, where the tag fires on a button click or every call is counted as a win. The number looks fine until you ask it which leads became paying jobs.
03 Cost per lead, by trade
What a lead costs,
clean accounts only.
The seven accounts with conversion tracking I trust, in the last 90 days. A lead here is a tracked call or form, not a booked job. The booked job is the number that actually matters, and it lives in the CRM, not the ad platform.
| Trade | Cost per lead |
|---|---|
| Septic | $72 |
| Lawn & tree care | $88 |
| Plumbing | $130 |
| Painting | $138 |
| Fencing | $175 |
| Landscaping | $254 |
| Roofing | $258 |
| Median | $138 |
The order tracks job value, not effort. Septic and lawn care leads are cheap because the jobs are smaller and the searches are more specific. Roofing and landscaping leads cost more because the jobs are larger, the consideration is longer, and more of the clicks are early-stage shopping. A $258 roofing lead is not expensive if the roof is $15,000 and you close one in four.
04 What it means
The dashboard number
is not the lead.
Put the two findings together. Cost per lead varies three to four times across the trades, so a benchmark borrowed from another industry is useless. And in a third of real accounts, the conversion number is inflated or missing, so even the in-platform cost per lead is often wrong. The dashboard is measuring button clicks and phone rings. Your crew gets paid for booked jobs.
The fix is the same one I run on every service account. Make a qualified lead the one primary conversion, track the calls properly, and feed the booked job from your CRM back into Google as an offline conversion. Then Smart Bidding optimizes toward the clicks that fill the calendar, not the ones that ring the phone and go nowhere. The full architecture is the Lead Quality Stack, and the question-by-question diagnosis is the Lead Quality library.
If you run a home-services shop, the home-services page covers how I rebuild an account to this standard, trade by trade.
Want this run on your account?
I will tell you which
of your numbers are real.
Thirty minutes on the phone. I look at your conversion setup, your call tracking, and how your booked jobs flow back to Google before the call. You walk out knowing which leads to trust and what to fix first.