What a home-services lead costs on Google Ads.
By Conner Crowe · Q3 2026, July to September · Published , data refreshed , wording updated · Download the data · Next edition January 2027
These are 11 home-services businesses whose Google Ads I manage, one per trade. It shows what they paid from 1 July to 30 September 2026, not what Google costs every business in these trades. A lead here is a call, form or Local Services lead that Google Ads counted, not a booked job. N next to a figure is how many businesses it comes from.
The short version
- I looked at the Google Ads of 11 home-services businesses I manage, one per trade, from July to September 2026.
- For the 9 whose lead counts I trust, a lead cost the typical business $163 (N = 9). With one business per trade, a septic company paid $98, a plumber $196 and a roofer $480 (N = 1 each).
- A lead isn’t a booked job. Divide the cost by the share of leads you book to get what a job costs you.
- In all 4 businesses that ran both, a lead from regular search ads cost 3.8 to 14.3 times as much as a lead from Local Services Ads (N = 4), though the two count leads differently.
- 2 of the 11 had lead counts too broken to price a lead (N = 11). In the 2 that stopped counting taps on their phone number as leads, the reported cost per lead rose about six and about elevenfold at the same click price (N = 2). The higher figure is the more honest one.
- $163 typical cost per lead, businesses with tracking I trust (N = 9)
- $13.27 typical cost per click (N = 11)
- 10.3% of clicks became leads, typical business (N = 9)
- 2 of 11 had tracking too broken to say what a lead cost (N = 11)
All four: home-services businesses whose Google Ads I manage, 1 July to 30 September 2026.
Cost per lead varied
widely by trade.
Across the 8 trades shown, one business each, a lead cost from $98 for a septic company to $480 for a roofer (N = 8). The typical business of the nine whose lead counts I trust paid $163, and the middle half of them paid between $112 and $196 (N = 9). Each row is one business, so read it as one example, not the going rate for that trade.
Cost per lead by trade, July to September 2026
One business per trade. The tick marks the July 2026 edition, which this one replaces.
| Trade | Cost per lead, Jul to Sep 2026 | July 2026 edition (replaced) | Businesses, N (now / July) |
|---|---|---|---|
| Septic | $98 | $84 | 1 / 1 |
| Lawn and tree care | $112 | $80 | 1 / 1 |
| Painting | $116 | $123 | 1 / 1 |
| Fencing | $163 | $164 | 1 / 1 |
| Pest control | $171 | Excluded: tracking | 1 / 1 |
| Plumbing | $196 | $128 | 1 / 1 |
| HVAC | $438 | Excluded: tracking | 1 / 1 |
| Roofing | $480 | $309 | 1 / 1 |
| Paving | Excluded: tracking | Excluded: tracking | 1 / 1 |
| General contractingc | Excluded: tracking | Excluded: tracking (114% conv. rate) | 1 / 1 |
| Landscaping | Not in Q3 sample | $226 | 0 / 1 |
| One other July account (trade withheld) | Not in Q3 sample | Excluded: tracking | 0 / 1 |
| Typical | $163 | $128 | 9 / 7 |
| Middle half | $112 to $196 | $104 to $195 | 9 / 7 |
| Full range | Not shown | $80 to $309 | 9 / 7 |
- Eight of the nine have a row. The ninth counts in the typical figure and the middle half but has no row, because it belongs to a client I name elsewhere. That’s why the full range isn’t shown: with eight rows on the page, it would give that account’s number back.
- One account in the table still counts some click-to-call taps as calls, so its true cost per lead is higher than shown (N = 1).
- c General contracting: the July 2026 edition carried this account under a different trade label, as the account at a 114% conversion rate. Its tracking still counts page scrolls and form starts (N = 1).
- July values are the July 2026 edition as published: a 90-day window ending 17 July 2026, 12 accounts. The windows overlap by 17 days, so read any single row’s change as a direction, not a precise move.
Roofing moved most between editions, from $309 to $480, and plumbing went from $128 to $196 (N = 1 each). Fencing ($164 to $163) and painting ($123 to $116) held (N = 1 each).
To turn any row into a cost per booked job, divide it by your close rate. I walk through that math in cost per booked job for home services.
In 4 accounts, Local Services
leads cost less.
Four businesses ran both Local Services Ads, Google’s pay-per-lead ads for local service businesses, and regular search ads (N = 4). In every one, a search lead cost more: 3.8 to 14.3 times as much. Three of the four gave Local Services 6% of their spend or less.
| Trade | Local Services cost per lead | Search cost per lead | Search costs this many times more | Share of spend on Local Services | N |
|---|---|---|---|---|---|
| Painting | $32 | $138 | 4.4x | 6% | 1 |
| Pest control | $46 | $174 | 3.8x | 1% | 1 |
| Roofing | $39 | $562 | 14.3x | 1% | 1 |
Three of the four have a row here: painting, pest control and roofing. The fourth counts in the ranges but has no row, and I don’t give a typical figure, because with three rows on the page a typical figure for four would give the fourth back. These costs are per channel, so they differ from the account-wide figures above.
The two don’t count leads the same way, and nothing here adjusts for it. A Local Services lead is a call or message Google charges for. A search lead is whatever the account counts. How many Local Services leads arrive also depends on how many Google sends, so more budget doesn’t always buy more of them. Even so, search leads cost more in all four.
What a click costs, and
how many become leads.
Cost per lead comes from two numbers: what a click costs, and how many clicks turn into leads. The typical business paid $13.27 a click (N = 11) and turned 10.3% of clicks into leads (N = 9). The middle half paid between $11.93 and $19.48 a click. Each typical figure is a different middle business, so dividing one by the other won’t give the typical cost per lead.
- Most count only phone calls. 6 of 11 counted phone calls as their only lead (N = 11).
- Mobile converted higher in 8 of 9. Mobile turned clicks into leads at a higher rate than desktop in 8 of 9 businesses with clean tracking, fencing the exception (N = 9). Part of that is what gets counted: 6 of 11 count only phone calls, and calls come mostly from phones. The typical business got 78% of its clicks on mobile (N = 11).
- Budget, more than rank, limited how often ads showed. Impression share is how often the ad showed when it could have. The typical business showed on 36% of the searches it was eligible for (N = 11). In 7 of 11 accounts, more of the missed searches were lost to budget than to Ad Rank, Google’s score from the bid and the ad’s quality (N = 11).
- Some spend buys clicks that never become leads. Google shows the searches behind only part of the spend. Of the search spend it showed, the typical business with clean tracking spent 27% on searches that got 2 or more clicks and no lead (N = 9).
The full table: typical, middle half and range for each measure
Each business counts once. A cell that says not shown would give back the number of the account without a row.
| Measure, Jul to Sep 2026 | Typical | Middle half | Range | N |
|---|---|---|---|---|
| Cost per click | $13.27 | $11.93 to $19.48 | $7.73 to $34.14 | 11 |
| Conversion rate, clicks to leads | 10.3% | 8.1% to 11.8% | 6.3% to 13.2% | 9 |
| Search impression share | 36% | 26% to 45% | 21% to 52% | 11 |
| Mobile share of clicks | 78% | Not shown | 60% to 90% | 11 |
| Visible search-term spend on terms with 2+ clicks and no conversion (clean-tracking accounts) | 27% | 24% to 34% | Not shown | 9 |
Honest tracking made
cost per lead go up.
Google Ads only knows what a lead is if the account tells it. Many accounts tell it the wrong thing: a page scroll, or a tap on the phone number even when no call connects. That makes the lead count too high and the cost per lead look too low. Others count too little. 2 of the 11 got it so wrong I couldn’t price a lead, and two more passed with a caveat (N = 11). These are normal accounts, set up the normal way.
-
1
tracked lead on the paving account in the whole quarter. Only calls from ads were counted. No website call or form was counted, so the dashboard couldn’t say what a lead cost. N = 1.
-
114%
conversion rate on the general contracting account. It counted a 50% page scroll and starting to type in a form as conversions, so it logged more conversions than clicks and no completed leads. N = 1.
-
39%
of one account’s conversions were click-to-call taps, a tap on the number rather than a connected call. It stays in the table, and its true cost per lead is higher than shown. N = 1.
In the July 2026 edition, three accounts were over-counting, at 67%, 69% and 114% conversion rates. In the first two, a tap on the phone number counted as a call. Both have since stopped counting most of those taps, and now report 6.3% and 11.0% (N = 1 each). One still counts some taps (the 39% above). The contractor is still broken. The typical clean account converts 10.3% (N = 9).
So the cost per lead Google reported for them went up. The clicks cost the same, but far fewer of them were counted as leads. Fixing the count multiplies the reported cost per lead by about eleven for one account and about six for the other (N = 1 each). The higher figure is the real one, and the one you can bid against. The math: cost per lead is cost per click divided by the share of clicks that become leads. At 67%, a lead costs about one and a half clicks. At 6.3%, it costs about sixteen.
In July, 4 of 12 accounts were broken (N = 12). Now it’s 2 of 11 (N = 11).
If you run a home-services
business, here’s what I’d do.
- Compare yourself with your own trade. Cost per lead ran from $98 to $480 across eight trades in one quarter (N = 8), so a figure borrowed from another trade tells you little. Then divide by your close rate to get cost per booked job.
- Check what your account counts as a lead. A page scroll or a tap on the phone number isn’t a lead. Count one thing: a qualified lead. The Lead Quality Stack is how I set that up.
- Send booked jobs back to Google. When the CRM reports which leads became jobs, Google’s bidding can chase the clicks that fill the calendar. The Lead Quality library covers it question by question.
- If cost per lead is climbing, split it in two. Find out whether the click got dearer or fewer clicks became leads. This is how I split it.
Paid search now reaches inside Google’s AI answers, too. In a one-day test across five cities, plumber, electrician, pest control, house cleaning and moving questions drew an ad inside the AI Mode answer in all five. The home-services page covers how I rebuild an account, trade by trade.
Since July, the typical lead
went from $128 to $163.
Most of that rise is a change in which businesses are in the figure, not prices. The July 2026 edition stays on the page, marked as replaced, in the cost-by-trade table.
- Fewer businesses: 11, down from 12. Landscaping and one other July account aren’t in this edition. One account is new. It belongs to a client I name elsewhere on this site, so it counts in every typical figure here but doesn’t get a row. The account July listed under a different trade label is labelled general contracting now.
- A fixed quarter: 1 July to 30 September 2026, replacing the 90-day window that ended 17 July. The two windows overlap by 17 days.
- Typical cost per lead: $163 (N = 9), up from $128 (N = 7). Most of that is which businesses are in it. Two accounts passed the tracking check for the first time and joined the figure (N = 2). On the six trades with a figure in both editions, the typical cost went from $126 to $139 (N = 6).
- Broken tracking: 2 of 11 accounts, down from 4 of 12. The general contracting account is broken in both editions.
- Withdrawn: the July total of spend I managed. The benchmarks standard publishes prices and rates, never spend, and that figure broke it.
- Added: an N on every row, the middle half, the Local Services comparison, figures you can quote, a CSV and a citation.
How I measured, and
how sure to be.
- Where it comes from. Google Ads API reads on accounts I have managed, never named, with read-only access. No survey, vendor report or industry average is mixed in.
- The window. Fixed: 1 July to 30 September 2026. Editions follow calendar quarters from here on, and old figures stay on the page, marked as replaced.
- Who’s in it. Every home-services account I managed with at least $300 of spend in the window. Eleven qualified, one per trade. Accounts below that floor or closed before the window aren’t in the sample.
- Small numbers. One business per trade means a row is one example, not a rule. The typical figure across nine is steadier, but it’s still nine businesses, and a different nine could land somewhere else.
- Accounts without a row. Some accounts belong to clients I name elsewhere on this site. They count in every typical figure and total here, but I don’t show them as their own row, because a row would publish a named client’s numbers. That’s one of the 11 here, so the tables and the CSV have 10 rows. Where a figure would let you work that account’s number back out from the rows shown, the table says not shown.
- What counts as a lead. The account’s primary conversion actions in Google Ads: calls from ads, website calls, form submissions, and Local Services phone and message leads. Secondary actions such as page views are ignored. In two accounts, a tap or click on the phone number also counts, and one of them is flagged in the CSV. Booked jobs aren’t in this data.
- Cost per lead. Account-wide: the account’s spend divided by its leads, across every campaign type, Local Services Ads included where the account runs them. That matches the July method.
- Who gets left out. An account comes out of the cost-per-lead figures when its tracking is broken: near-zero leads for the quarter (paving, one), or counted actions that aren’t leads (general contracting, scroll and form-start events). Left-out accounts still count in the tracking findings, which use N = 11. An account that partly counts taps as calls stays in (HVAC, footnoted, and one other account to a smaller degree).
- Typical and middle half. Each business counts once, however much it spends. The typical figure is the median. The middle half is the 25th to 75th percentile by linear interpolation, the same method as a spreadsheet’s QUARTILE.INC, and is shown where N is 4 or more.
- Rounding. Cost per lead to whole dollars, cost per click to the cent, conversion rates to one decimal place, shares to whole percents. The CSV carries two decimals.
- Comparing to July. The July 2026 edition used a 90-day window ending 17 July 2026 and 12 accounts. Rows are matched by account, and one account carried a different trade label in July (note c). With N = 1 per row, a single row’s change is a direction, and the typical figure is the one to compare.
- Published under the benchmark rules: no client named, no spend, N next to every figure.
Words used here
- Lead
- A call, form or Local Services lead the account counts as a conversion. Not a booked job: that number lives in the business’s CRM.
- Cost per lead (CPL)
- What the business spent on Google Ads divided by the leads it got.
- Cost per click (CPC)
- What a business pays Google each time someone clicks its ad.
- Conversion rate
- Out of every 100 clicks, how many became a lead. 10.3% means about 10 in 100.
- Impression share
- Out of all the searches an ad could have shown on, the share where it did show.
- Local Services Ads
- Google’s pay-per-lead ads for local service businesses, usually shown at the top of results.
- Typical
- The middle business when you line them all up from lowest to highest. With an even number, it’s the average of the two in the middle. Statisticians call it the median.
- Middle half
- Line the businesses up from lowest to highest and drop the bottom quarter and the top quarter. The middle half is the range of what’s left. It shows how spread out they were.
- Close rate
- The share of leads that become booked jobs.
- N
- How many businesses a number is based on. N = 1 means one business.
Figures you can quote.
Each line carries its scope, so it stays accurate when it’s copied on its own. If you use one, link this page so readers can check it.
- In 9 home-services Google Ads accounts managed by Conner Crowe with clean conversion tracking, the median cost per lead was $163 in Q3 2026, 1 July to 30 September. Managed accounts, not a market measure (N = 9).
- Across the 8 home-services trades shown as rows in Conner Crowe’s managed Google Ads accounts, cost per lead ran from $98 (septic) to $480 (roofing) in Q3 2026, one account per trade (N = 8).
- 2 of 11 home-services Google Ads accounts managed by Conner Crowe had conversion tracking too broken to report a cost per lead in Q3 2026 (N = 11).
- Two home-services Google Ads accounts managed by Conner Crowe that reported 67% and 69% conversion rates in the July 2026 edition reported 6.3% and 11.0% in Q3 2026, once taps on their phone numbers mostly stopped counting as leads (N = 2).
- In all 4 home-services Google Ads accounts managed by Conner Crowe that ran both Local Services Ads and Search in Q3 2026, Search leads cost 3.8 to 14.3 times as much as Local Services leads (N = 4).
- The median cost per click across 11 home-services Google Ads accounts managed by Conner Crowe was $13.27 in Q3 2026 (N = 11). The roofing account paid the most, $34.14 (N = 1).
- 6 of 11 home-services Google Ads accounts managed by Conner Crowe counted phone calls as their only primary conversion in Q3 2026 (N = 11).
- Mobile converted at a higher rate than desktop in 8 of 9 home-services Google Ads accounts managed by Conner Crowe with clean tracking in Q3 2026 (N = 9).
How to cite
Crowe, Conner. "Home Services Google Ads: Cost Per Lead by Trade." Q3 2026 edition. Published 29 June 2026, refreshed 3 October 2026. https://connercrowe.com/home-services-paid-search-benchmarks/
If you quote a July figure, label it the July 2026 edition. Data licensed CC BY 4.0, attribution to Conner Crowe with a link.
Download the Q3 2026 data as CSV. Ten anonymized rows, one for each account I can show: code, trade, N, tracking status, cost per lead, the superseded July cost per lead, cost per click, conversion rate, impression share, mobile click share, Local Services share of spend, zero-conversion search-term share, Local Services and Search cost per lead, whether mobile beat desktop, and the window. No spend.
Common questions.
How much does a home services lead cost on Google Ads?
What is a good cost per lead for roofing on Google Ads?
What conversion rate should a home services Google Ads account have?
Are Local Services Ads cheaper than Google Search ads for home services?
What is the average cost per click for home services Google Ads?
Why did my cost per lead go up after I fixed conversion tracking?
More first-party benchmarks
I’ll tell you which
of your numbers are real.
Add me as a read-only user and I'll write the audit myself: which leads to trust and whether booked jobs reach Google, 5 to 7 pages, back within 24 hours of access. No call needed. If you'd rather talk first, a call is open too.