Google Ads for window cleaning companies, with real budgets

By Steve Meitler · 2026-07-15 · 6 min read

What to spend, which campaigns book jobs, the negative keywords that stop the waste, and the cost per lead math a $260 average ticket can actually carry.

Google Ads punishes low-ticket trades that run it carelessly. At a $260 average job with maybe 45% gross margin after crew and materials, you have around $117 of gross profit to work with. If you pay $70 for a lead and close half of them, your acquisition cost is $140 and you just lost money on a first-time customer. Run the same account with $28 leads and the job pays for itself the day it is cleaned.

The difference is almost never the bidding. It is match types, negatives, geography, and whether the phone gets answered. Here is how the account should be built.

Budget: what is enough to learn anything

Under $800 a month, Google Ads for window cleaning is a coin flip. You get too few clicks to know what worked and the algorithm never leaves its guessing phase.

Realistic bands:

• $800 to $1,200 a month: one or two cities, search only, exact and phrase match. Expect 25 to 45 leads.
• $1,200 to $2,000 a month: a metro, plus a second campaign for gutters or pressure washing in the shoulder seasons.
• $2,000 to $2,500 a month: multi-city, plus Local Services Ads running alongside.

Weight the spend by season. Push 60% of the annual budget into March through June and the four weeks before Thanksgiving, and pull back hard in January. The seasonality guide has the month-by-month version.

The campaigns that book jobs

Search, exact and phrase only. This is the core and it should be most of the money. Build tight ad groups around how people actually type it:

• "window cleaning [city]"
• "window washing [city]"
• "window cleaners near me"
• "residential window cleaning"
• "commercial window cleaning"
• "storefront window cleaning"

Six ad groups, each with its own ads, each pointing at a matching page. An ad group for commercial should not land on your residential page. Bid higher on the commercial and storefront terms even though they convert less often, because a storefront route is recurring revenue at 12 to 26 visits a year, not a one-off $260.

Local Services Ads. Where Google runs them for cleaning categories, they sit above everything and you pay per lead rather than per click, often $15 to $45. They require a background check and proof of insurance, which takes a week or two. If you can pass the check, do it, and note that the Google Guaranteed badge is a trust marker in a trade where a stranger climbs a ladder over your car.

Performance Max, carefully. PMax will spend your money on Display and YouTube placements that produce cheap clicks and no calls. If you run it, cap it at 20% of the budget, exclude all placements you can, feed it real conversion data and never let it be the only campaign. For most window cleaning companies under $2,000 a month, skip it entirely and put the money in search.

Where the money leaks

Broad match. Turn it off. "Window cleaning" on broad match will match to window replacement, window tinting, gutter guards, and car detailing. Exact and phrase, with a short broad-match test only after you have 90 days of conversion data.

The negative keyword list. Build it before you turn the account on. Start here:

window replacement, window repair, window tinting, window film, windshield, window installation, blinds, curtains, glass repair, glass replacement, screen repair, window air conditioner, jobs, hiring, salary, how to, DIY, free, cheap, franchise, equipment, supplies, wagner, unger

Review the search terms report every Monday for the first month and add to that list. Expect to add 20 to 40 terms in the first 30 days. Skipping this step is the single most common reason a window cleaning account burns $1,000 with nothing to show.

Geography. Set targeting to "presence" and not "presence or interest," or you will pay for clicks from people in another state who searched your city once. Set a radius that matches your actual drive time, usually 15 to 25 miles, and exclude the parts of it where you do not want to send a crew.

Ad schedule. Run ads when someone answers. If the office is 7 a.m. to 5 p.m. Monday through Saturday, do not bid at 11 p.m. A window cleaning search at midnight is a price shopper, and you will pay $6 for the click and get a voicemail. If you do run wider hours, back it with text-back, covered in the missed calls guide.

The cost per lead math

Realistic cost per lead in this trade runs $18 to $60 depending on the metro. Denver, Phoenix and the California coastal cities are at the top; smaller Midwest markets sit at the bottom.

Do the arithmetic on your own numbers rather than trusting a benchmark:

• $1,500 spend, 40 leads, $37.50 per lead
• 40 leads closed at 45% is 18 jobs
• 18 jobs at $260 is $4,680 in revenue
• Acquisition cost per job: $83, against roughly $117 of gross profit

That works, but barely, and only if the customer comes back. The number that saves this account is the repeat. Most residential window cleaning customers rebook every six to twelve months, and a customer who books twice a year for three years is worth $1,560. The ad has to be judged on the second job, not the first. Put your own numbers into the calculator to see the same math on your ticket.

Two levers make it comfortable. Raise the ticket by bundling: offering a gutter clear or a driveway wash on the same visit typically adds $120 to $200 without another drive. And close more of what you already pay for, which is a phone answering problem far more often than an ad problem.

Tracking, or you are guessing

Install conversion tracking before the first click. At minimum:

1. A Google Ads call-tracking number on the site, with calls over 60 seconds counted as conversions.
2. Form submissions as a separate conversion.
3. Call-from-ad conversions on the ad extension.
4. Call recording turned on, so you can hear why the $37 lead did not book.

Then look at the recordings monthly. Owners are usually shocked. The problem is often a crew member answering on a ladder, or a quote never sent, and no keyword change fixes that.

Ads work best on a profile and site that are already in decent shape. If your listing sits near the median of 7 reviews and 32% of the trade has no website at all, spend the first month on the profile and reviews and start ads in month two. The same click converts far better against 60 reviews than against 6.

Frequently asked

How long before I know if it is working? Six weeks at a real budget. The first two weeks are mostly negatives and search terms. Judge it on booked jobs and cost per booked job, not on clicks or impressions, and do not restructure the account weekly.

Should I bid on my own company name? Yes, if a competitor is bidding on it, which you can check by searching. It costs a few dollars a month because your quality score on your own name is near perfect, and it stops a competitor from buying the customer you already earned.

Is Bing worth it for window cleaning? Small volume, but cheap. Clicks often run 30 to 50% less than Google and the audience skews older and homeowning. Import the campaign once Google is stable and give it $150 a month.

Account builds and management are part of the free 14-day trial, and you keep the account either way. Text or call (385) 832-6175.