Does a carpet cleaner need a big website to reach the map pack top three?

Carpet cleaning owners are told that the map pack goes to the companies with a big, busy website: a blog, a page for every suburb, a service page for every stain. A cleaner with a ten page site, the story goes, is wasting its time.

We tested that belief against the sites themselves. On October 6, 2026 we captured the Google Maps top 20 for “carpet cleaning” in 22 US cities, 11 large metros and 11 small cities. That was 433 results from 22 searches. After screening, 359 were carpet cleaning businesses, and we read the sitemap of every website we could reach.

The short answer: the independents in the top three do have somewhat bigger sites than the ones at the bottom of the list, but “big” is a few dozen pages, not hundreds. And the biggest sites in the sample belong to national brands, which would have told a completely different story if we had left them in.

The number to write down first

The median independent carpet cleaner in the map pack has a website of 25 pages. A quarter of them have 10 or fewer, and a quarter have 79 or more. That is from 174 independents whose sitemap we could read.

Here is the same measure by position on the list, independents only:

Maps position	Independents	Sitemaps read	Median pages	Median reviews
1 to 3	41	29	36	178
4 to 7	58	34	40	122
8 to 12	76	45	23	53
13 to 20	111	66	22	65

To place yourself, open your own sitemap, count the URLs, and compare. If you have around 36 pages you look like the typical independent in the top three. If you have 22 you look like the typical independent at the bottom. Neither number is large.

Where the belief breaks

Across the 29 top-three independents whose site we could read, 10 had fewer than 25 pages, the median for all independents, and 4 had fewer than 10. A carpet cleaner with a ten page site is not shut out of the top three. It is in there in four of 29 cases.

The relationship between position and site size is real but weak. Among independents, a higher position goes with a bigger site at a rank correlation of -0.16 across 174 sites, which is statistically significant and small. Nothing about the bottom of the list says the owners there need a blog. Having any website at all does not sort the list either: 242 of 286 independents have their own site, 37 of the 41 in the top three do, and the correlation between position and having one is -0.022, no meaningful relationship.

Metros and small cities

The shape holds in both kinds of market, with different levels.

Market	Median pages, top three	Median pages, 13 to 20	Sitemaps read
Large metros	45	27	124
Small cities	25	17	50

In the 11 large metros, independents in the top three carry 45 pages at the median, against 27 for positions 13 to 20, and the correlation holds up (-0.181 across 124 sites). In the 11 small cities the top three carry 25 against 17, and with only 50 readable sites the correlation is -0.095, not significant. That does not mean no effect. It means 50 sites cannot detect an effect that small.

In small cities, 6 of the 12 top-three independents we could read had fewer than 25 pages and 3 had fewer than 10. In metros, 4 of 17 had fewer than 25 and 1 had fewer than 10.

Reviews, not pages, are the steeper ladder in small cities. Independents in the small-city top three carry 136 reviews at the median against 14 at positions 13 to 20, and the correlation between position and reviews is -0.469 across 110 listings. If you have to choose where to spend a month, the data points at the review count first.

How fresh is the site

Freshness is the other place the top three look different. Counting sitemap URLs updated in the past year, independents in the top three have a median of 23 against 5 at positions 13 to 20 (the middle two bands sit at 6 and 11). The rank correlation is -0.137 across 167 sites and is not significant, so treat it as a lead, not a result. The top three look like small sites that are touched often, not large sites left alone.

What the brands do to the average

Of 359 listings, 73 are locations of a national or regional brand, 20 percent of the list and 23 of the 64 top-three slots. One franchise has 21 locations across 19 of our 22 cities, and 10 of them sit in the top three. Their sitemaps are the brand’s, not the local office’s: the median brand site is 562 pages and the largest is 60,203, the corporate site of a restoration franchise. The median independent is 25. Averaging them together would have told an owner with a 25 page site that the benchmark was far out of reach, when the typical independent in the top three has 36.

The other things we checked

We scored each readable homepage out of six: a contact form, a pricing link, a free estimate offer, a satisfaction guarantee, pet stain or odor language, and an insurance mention. The median site has 2 of 6, 42 percent have 3 or more, and 4 percent have 5. The top three have a median of 3. Pricing links are rare, 16 percent of top-three sites among the readable ones. None of the six sorts the list on its own.

Limits

• Range restriction. We measured only businesses that already appear in Google’s top 20. This says what the sites of ranking cleaners look like. It cannot say what separates them from cleaners who do not appear.
• Coverage. We read a sitemap for 217 of 359 sites, 60 percent, and 174 of the 286 independents. Coverage was 69 percent in metros and 48 percent in small cities. A site with no readable sitemap is missing, not counted as small, and the missing ones are more likely to be small, which would make the real medians lower.
• Pages are a proxy. A sitemap count measures how many addresses a site publishes, not whether they are good. Reviews, proximity, and the business profile itself also drive map position, and bigger sites belong to businesses that have usually also earned more reviews. We did not isolate the website’s own effect.
• One day. Maps results move with the searcher’s location and the day. We searched once, from no specific neighborhood.
• Screening. We excluded 64 results: house cleaners, dry cleaners, janitorial companies, carpet stores, restoration firms and similar neighbors, each with a reason in the dataset. Franchise detection works from host names and brand names, and a small brand we did not recognize could be counted as independent.
• Not measured. Site speed, design, backlinks and how the business profile is managed.

The dataset behind every figure here is linked below, one row per listing, with the reason attached to every exclusion.