We count both trades every week, the same way. 216 bath companies and 153 window companies, on the same Monday. The ages, the buttons and the franchise gap all run differently.
You sell an in-home estimate. So does the window company down the road. You would think the ads look alike.
We count every bath company and every window company advertising on Meta in America every week. On 14 September 2026 that was 216 bath companies running 4,389 live ads, and 153 window companies running 2,623.
Same country, same week, same method. Here is where they part.
| Week of 14 September 2026 | Bath | Windows |
|---|---|---|
| Companies advertising | 216 | 153 |
| Live ads | 4,389 | 2,623 |
| Promise a free estimate | 155 | 95 |
| Put "Book now" on the ad | 11 | 14 |
| "Learn more" share of live ads | 61% | 48% |
| Median company's ads, by age | 19 days | 37 days |
| Top quarter, by age | 49 days | 103 days |
| Oldest live ad | 712 days | 2,508 days |
Base: every verified advertiser in each trade, each page queried directly, 14 September 2026.
This is the biggest gap on the board, and it is not close.
The median bath company's live ads are 19 days old. The median window company's are 37.
Go to the top quarter and it widens. 49 days in bath, 103 in windows.
The oldest live bath ad has been running 712 days. The oldest window ad, 2,508. That is nearly seven years, and one opening line in the window trade has been live 2,663 days.
Now the part we got wrong ourselves, until we checked.
An age is not a lifetime. Every ad we can measure is one that has not stopped yet. The ages above are the ages of survivors.
A trade that launches many new ads has young ads. A trade that launches few has old ones. That holds whatever happens to any single ad.
So "window ads are older" and "window ads last longer" are two different sentences. We counted the first.
The second needs the same ad found twice, a fortnight apart. For bath we have it. Of the 2,045 bath ads live on 4 September, 1,529 were still running on the 14th. That is 75%.
We are not giving you the window number this week. Our window sweep came back with less than the sweep before it. An ad we failed to collect looks like an ad that stopped. We will not guess which.
So the useful read is not which trade's ads last longer. We do not know that yet. It is this. The median window company has ads a month older than the median bath company's. And nobody in either trade can tell you which of those ads put a customer in a chair.
In bath, a page under a franchise system runs a median of 6 live ads. An independent page runs 18.
In windows it flips. A franchise page runs 18. An independent runs 11.
| Median live ads per page | Bath | Windows |
|---|---|---|
| Under a franchise system | 6 | 18 |
| Not a franchise | 18 | 11 |
Base: 224 bath pages and 174 window pages, same week.
Half the bath trade sits under a system. Under a third of the window trade does.
So in bath the independent out-runs the unit three to one. In windows the system is the one doing the volume, and the independent is the quiet one.
If you run a bath unit, your system is the reason you run six ads. If you run a window shop on your own, your competitor under a brand is running nearly twice what you are.
11 of 216 bath companies put "Book now" on the ad. 14 of 153 window companies do.
That is 5% against 9%. Both are small. Neither is good.
"Learn more" carries 61% of bath ads and 48% of window ads. In both trades the most common button sends a homeowner off to read something instead of to a calendar.
This week we also read the picture on every ad in both trades. 2,463 bath pictures and 1,604 window ones.
| What the picture shows | Bath | Windows |
|---|---|---|
| A real job on site | 60% | 42% |
| The crew | 10% | 22% |
| Stock imagery | 4% | 12% |
| A before and after | 10% | 6% |
| A price | 25% | 30% |
| Readable with the sound off | 85% | 81% |
Base: 2,463 bath pictures and 1,604 window pictures, read the same week by the same method.
A bathroom changes in a way a camera can hold. A window mostly does not, so the window trade points the camera at the people instead.
That is also why bath runs before and after twice as often, and why windows leans on stock three times as much.
The story types split the same way. Bath leads with a walkthrough, 28% of its pictures. Windows leads with someone talking to camera at 24%, then a list at 23%.
The most repeated opening line in bath runs in 125 versions with the county swapped. In windows the most repeated line runs in 36.
Bath makes one idea many times. Windows makes one idea and leaves it up for years.
Both are ways of not starting from scratch every month. Most owners in both trades do neither.
Nothing here is spend or results. A long-running ad is not proof of a sold job, and a company with 40 ads is not beating one with 6.
The two rosters are also different sizes, so read the shares and not the totals. We publish the method, the coverage and every row, and we correct them in public when we get one wrong.
Find your own trade's line in the table above. Then check your account against it.
If your ads are younger than your trade's median, you are paying to rebuild more often than your competitors. If they are much older, nobody has looked at them.
And in both trades the same thing is true. The button almost everyone uses sends her away to read. 11 bath companies and 14 window ones point her at a calendar instead.
Lovro Lucic, on US hours from Zagreb. Your page, your number and your calendar stay in your time zone. Call +1 (650) 643-4497, or pick a time on the calendar.
Want to see how your own ads read?
Send us your Facebook page and within a day a person sends back what you are running, the button on it, whether your page can book a time, and your offer next to your competitors'. Free.