ResultsArizona Window Washing Pros
We Were Close to Closing the Business. Then We Changed Our Strategy.
How a market pivot helped turn around Arizona Window Washing Pros
- Business
- Arizona Window Washing Pros
- Industry
- Window cleaning
- Market
- Arizona
- Period
- January 1 – October 19, 2026
Sometimes the Problem Isn't Your Service. It's the Market You're Trying to Sell It In.
The business
We started Arizona Window Washing Pros in January 2026. I own and operate the company. That matters here, because the advertising budget was not a client's test spend. It was money the business had to earn back after labor, supplies, fuel, and insurance.
The service was professional window cleaning. The open question was whether we were trying to sell it in a market that could support the cost of getting the lead.
What wasn't working
The early advertising was not profitable. Cost per lead sat around $60. We closed roughly 30% of the leads, and the average ticket was about $200.
At that closing rate, acquiring a customer cost about $200 before labor, supplies, fuel, insurance, and the rest of the operating costs. The lead was consuming the job.
By June, the advertising economics were unsustainable. Spending more inside the same approach would have made the problem larger, not smaller.
The strategy
Sometimes you don't need a better ad. You need a better strategy. In June 2026 we stopped trying to win the same saturated market and looked for communities where professional window cleaning was underserved.
We focused advertising on Arizona's White Mountains, including Show Low, Pinetop-Lakeside, Snowflake, and surrounding communities. A service business does not have to compete in the largest market to succeed. Sometimes the better opportunity is where demand is real and the competition is thinner.
More ad spend isn't always the answer. The change was the market, the message, and who we were willing to drive to — not a promise that a bigger budget would fix a bad equation.
The results
The original period, January 1 through May 31, 2026, is still part of the story. Meta recorded 88 leads, $5,101.02 spent, and a $57.97 cost per lead. Those numbers are on the screenshot. I am not going to smooth them over.
From June 1 through October 19, 2026, the same ad account recorded 212 Meta leads, $6,232.45 in spend, a $29.40 cost per lead, and 175,901 impressions.
Meta cost per lead fell by about 49.3%. That figure is ($57.97 − $29.40) ÷ $57.97. It describes the lead, not the profit on a finished job.
We also logged about 150 Google-originated leads in our own business reporting for the later period. Those are separate from the 212 Meta leads. I do not have data that proves the two sources are entirely different people, so I do not add them into one customer total.
88
Meta leads before
January 1 – May 31, 2026
$57.97
Cost per lead before
$5,101.02 spent
212
Meta leads after
June 1 – October 19, 2026
$29.40
Cost per lead after
$6,232.45 spent
49.3%
Lower Meta cost per lead
Compared with January–May
~150
Google-originated leads
Internal reporting, not the Meta export
Reported closing rate moved from about 30% before the pivot to about 60% after it. Average project value was higher after the market change. I am not publishing a new ticket-size dollar figure, because I am not treating that increase as a verified revenue total.
Before the pivot, the business was financially strained. After it, lead volume was stronger and the operation was able to keep growing. That is the business outcome I will stand behind. It is not a claim of overnight wealth.
| Measure | January 1 – May 31, 2026 | June 1 – October 19, 2026 |
|---|---|---|
| Meta leads | 88 | 212 |
| Amount spent | $5,101.02 | $6,232.45 |
| Cost per lead | $57.97 | $29.40 |
| Impressions | 119,612 | 175,901 |
| Reported closing rate | About 30% | About 60% |
| Average ticket | About $200 | Higher after the pivot |
| Position of the business | Advertising cost about as much as the average job, before operating costs | Lower cost per lead, stronger lead volume, and a business that could keep operating |
Sometimes the Problem Isn't Your Service. It's the Market You're Trying to Sell It In.
The screenshots show both seasons on purpose. Data, and a willingness to leave a market that would not pay, mattered more than raising the budget.
Advertising screenshots
These are unedited exports. Click a preview to enlarge it. Nothing in the image has been altered.
What I take from it
The useful part of this story is that the first approach failed in a way we could measure. A $58 lead and a 30% close on a $200 ticket cannot carry a service business. Changing the creative without changing the market would have been activity, not a plan.
Leads don't pay the bills. Customers do. The later account shows more leads at a lower Meta cost per lead, and the business reported a higher closing rate on higher-value work. Those are different facts, and they only help if follow-up and production can handle the demand.
A Google Gemini response later described Arizona Window Washing Pros as a leading, or among the largest, window cleaning operators in the White Mountains. I am including the screenshot because it is part of the public record of that search. It is an AI-written description. It is not an audit, a market-share study, or proof that the company is definitively the largest in the region.
Methodology
- Both Meta totals are Leads (Form) — or the account's Leads column for this report — with the attribution setting shown in Ads Manager: 7-day click or 1-day view. A lead is a form submission, not a booked job or a collected payment.
- The 49.3% figure compares Meta cost per lead between the two exported periods. It is not a profit margin and it does not include operating costs.
- The closing rates, the $200 early ticket, and the note that tickets were higher after the pivot come from our own reporting. They are not cells in the screenshots.
- The approximate 150 Google leads are an internal count. They are not added to the 212 Meta leads.
- Impressions are included because both exports show them: 119,612 before and 175,901 after.