Behind the Agency Veil | No. 2
Many of the Google Local Services Ads Numbers You See Are Wrong. Here’s How to Tell.
What we learned when we traced every Local Services Ads cost figure we could find back to its source.
Brett Williamson, Founder & CEO of Ad Genius | adgenius.com
Ask what a Google lead should cost and you will get an answer fast. A blog post will give you one. An agency will give you one on the first call. An AI assistant will give you one in a single sentence, with total confidence.
The harder question is where that number came from.
When we built our national benchmark of Google Local Services Ads costs, we tried to answer that question for every figure we could find. Google Local Services Ads (GLSA) are the pay-per-lead ads at the top of Google’s local results, the ones with the Google Verified badge. They are also widely misquoted.
Many of the numbers you see did not survive the check. Some were accurate once and never updated. Some measured a different product entirely. Some traced back to no data at all. This piece walks through the eight ways a lead cost goes wrong, and gives you five questions to test any figure you are handed, including ours.
How We Checked
We reviewed more than 60 sources. For each figure, we went back to the original publication, not a summary of it, and asked four things: who measured it, when, across how many accounts, and what exactly was counted. A figure stayed in the benchmark only if we could answer all four. The full dataset then went through two separate verification passes before publication. As a final check, we compared the results with the Ad Genius client accounts we manage. Our clients’ numbers fall within the ranges we publish.
That standard is normal in research. It is less common in marketing, where a number can travel through several articles before anyone asks where it started. The benchmark’s methodology explains the process in full.
The Eight Ways a Number Goes Wrong
1. The Winter Snapshot
A figure you will see often is that a Local Services Ads lead costs about $53. That was true. It was the average across 888 businesses in February 2026.
February is off-season for many trades, and prices did not stay there. By June, the average lead cost $62.56 across 945 businesses, up 18%. The cost of each paying customer rose faster, from $233 to $302.33, up 30%. The February number still turns up in articles and answers published long after costs moved.
We are not exempt. Our own HVAC marketing costs article, published in July, quoted February’s HVAC figure of $51. By June the typical HVAC lead cost $59. We have updated the article, and we mention it here because a series about transparency has to apply to us too.
What it costs you: budget built on a winter price runs short in peak season, right when the phones should be busiest.
2. The Number From a Different Product
A set of figures that circulates widely puts HVAC leads at $52, cleaning at $28, and roofing at $71, credited to a Local Services Ads report covering more than 50,000 businesses.
We could not find that report. The only matching report we found measured regular Google Search ads, across 3,211 campaigns, not Local Services Ads. Search ads charge per click and Local Services Ads charge per lead, so the two are not interchangeable. In our benchmark, the typical HVAC Local Services Ads lead cost $59 in June 2026.
What it costs you: you end up judging one channel by another channel’s prices.
3. The Range With Nothing Behind It
Some of the most confident-sounding numbers are ranges: gutter leads at $20 to $60, foundation repair at $60 to $75, window cleaning at $20 to $50. You will find them repeated from site to site.
We followed each one back. Every trail hit one of three dead ends: a page that shows no data, a page that no longer loads, or no original source at all. The foundation range traces through one article to a page that could not be opened.
The one measured foundation repair account we found tells a different story. Over six months, it paid $131 to $153 per lead each month, about $140 on average across 181 leads. That is roughly double the circulating range. One account is not a benchmark. But it is measured data, and the range is not.
For these categories, the honest answer is that we found no verifiable benchmark. Our benchmark says so rather than filling the gap.
What it costs you: a range with no data behind it can make a healthy account look expensive, or an expensive one look normal.
4. The Big Sample Nobody Can See
A large number of accounts sounds like proof. One widely shared dataset claims nearly 10,000 accounts and puts the typical HVAC lead at $42 and plumbing at $38.
The accounts were blended from other published sources and the publisher’s own clients, with no way to see which figure came from where, which period it covers, or how leads were counted. A big sample you cannot inspect is not stronger than a small one you can.
Our benchmark’s June 2026 data covers 945 businesses and $13.1 million in spend, with its sample and period disclosed. It puts HVAC at $59 and plumbing at $71. It also shows plumbing costing more than HVAC at every point in the range, the opposite of what the blended dataset suggests.
What it costs you: a benchmark set too low makes a normal cost look like a problem, and can push you to cut a channel that is working.
5. The Lead Price Read as the Customer Price
A $59 lead sounds cheap. It is not a $59 customer.
A lead is a call, message, or booking request. Some of those become booked appointments, and some of those become paying jobs. At a 43.5% booking rate, the HVAC rate in March 2026 data, a $59 HVAC lead works out to about $136 per booked appointment, by our calculation. Only about one in five leads becomes a paying customer. In June 2026, the median HVAC account paid $281 per paying customer. The middle 80% of HVAC accounts paid anywhere from $129 to $716.
This mix-up is easy to make. We nearly made it ourselves: one range we collected turned out, on a second read of its source, to be a cost per booked job rather than a cost per lead. We caught it before publication.
What it costs you: judging the channel on lead price alone tells you nothing about whether it makes money.
6. The Right Number, Wrong Question
In August 2026, a study of 100 home service accounts found that about 60% of paid conversions, meaning the calls, messages, and form fills an ad produces, could not have been booked. It is a striking number, and easy to repeat as a Local Services Ads figure.
It isn’t one. That study graded conversions from all paid channels. A separate grading of Local Services Ads conversions alone, from roughly 100 contractors the same month, found 43% unbookable. The main causes were empty calls, automated phone menus, hang-ups, and requests for services the business does not offer. Of the conversions that could have been booked, 58% still were not.
Both numbers are real. They answer different questions. The 58% matters most to a contractor, because it is the part you can do something about.
What it costs you: blaming lead quality for a problem that is partly on your own phones.
7. The National Number Applied Locally
Even an accurate national average can mislead when it is applied to one market. In June 2026, the average lead cost $85 in the West and $62 in the Midwest and Northeast. Among states with at least 1,000 leads, costs ranged from $53 in Wisconsin to $89 in Washington.
A $70 lead can be a good result in Arizona, where the average was $84, and an expensive one in Michigan, where it was $57.
What it costs you: comparing yourself with the wrong market leads to the wrong decision in either direction.
8. The Rule That Changed
Some outdated information is not a number at all. It is a rule that no longer applies:
- “Google Guaranteed” protects your customers. Google replaced its Guaranteed, Screened, and License Verified badges with a single Google Verified badge on October 20, 2025. The money-back guarantee ended for services booked after December 7, 2025.
- You get credits for leads outside your job types or service area. Google no longer issues those credits, so your settings have to match the work you want.
- Missed calls cost nothing. From October 1, 2026, a missed call during business hours is charged as a lead when the caller stays on the line more than 20 seconds. Repeat calls from the same person within 15 days are charged once. Here is how missed-call charges work.
The platform itself is also changing. In August 2026, Google began moving Local Services Ads into Google Ads as a pay-per-lead version of Performance Max, its automated campaign type. We cover what the Performance Max migration means for your account separately.
What it costs you: advice written for the old rules can leave you paying for calls you never answer, or expecting protection that no longer exists.
Why Bad Numbers Spread
None of this requires anyone to be dishonest. It usually happens one step at a time.
An article cites another article, which cited a third. Each time a figure is repeated, it loses a little context: first the month, then the sample size, then what was being counted. A range is harder to quote than a single number, so ranges get rounded into one. And an old article that ranks well keeps getting read long after its numbers expire.
AI assistants add a new layer. More owners now ask an AI assistant what a lead should cost, and those tools answer from what they find published online. If the most visible number is a stale one, that is the number they repeat, stated with the same confidence as a fresh one. That is one reason we publish our benchmark with its dates, samples, and methods in plain view: so the next answer can be built on data that holds up.
Five Questions to Ask About Any Lead Cost
You do not need to audit 60 sources to protect yourself. Ask these five questions about any figure you are handed, whether it comes from a blog, an agency, an AI assistant, or us:
- Can it be traced to an actual measurement? A number with no measurement behind it is an opinion.
- When was it measured, and in what season? A February number and a June number are not the same number.
- How many accounts, and in which markets? Five accounts in one city tell you about those five accounts.
- Is it per lead, per booked job, or per paying customer? These can be several times apart.
- Is it the typical account, or an average pulled up by a few expensive ones? The median, the middle account, is usually the fairer comparison.
If a source cannot answer these, treat the number as a rough guess, however confidently it is stated.
What to Measure Instead
The most reliable benchmark is your own cost per paying customer: what you spend on Local Services Ads divided by the leads that became paying jobs. Compare it with your own trade, your own region, and the same season last year.
Then look at what happens after the phone rings. ServiceTitan’s data on its contractor customers found that top performers, the top 10% by revenue in their trade and market, book 62% of their inbound lead calls. Everyone else books 39%. At a $59 HVAC lead, that difference alone moves the cost per booked appointment from about $95 to about $151.
Our national benchmark has current figures by trade, region, state, and season, and our HVAC marketing cost breakdown walks through the math from lead to paying customer. These are the same ranges we see across the Ad Genius client accounts we manage.
Why We Publish This
Behind the Agency Veil exists because contractors deserve to see how marketing actually works, including the parts that make our industry look bad. Numbers are where trust is won or lost. A contractor who has been quoted three different lead costs by three different sources has every reason to doubt all of them.
So here is our commitment. Every figure we publish shows when it was measured, how many accounts it covers, and what was counted. When our numbers go stale, and they will, we will say so and update them, the same way we did with our own HVAC figure above.
If you want to know how your numbers compare with your trade, region, and season, that conversation starts at adgenius.com/schedule.
No pitch. A diagnosis first.
How to Cite This Research
Williamson, Brett. “Many of the Google Local Services Ads Numbers You See Are Wrong. Here’s How to Tell.” Ad Genius, Sep 29, 2026, adgenius.com/research/google-local-services-ads-statistics-that-dont-hold-up/.
Brett Williamson is the Founder and CEO of Ad Genius, a Phoenix-based digital marketing agency specializing in home service and professional service contractors. Ad Genius helps established contractors build the digital infrastructure needed to generate consistent, qualified inbound leads and stop competing on price because buyers can’t find them any other way. adgenius.com
