1,000 pages nobody wrote
3 min read
At its peak, QuickStaging's sitemap listed well over a thousand URLs, and almost none of them were written by hand. This is how they were built, what Google thought of them, and what I ended up deleting from the sitemap.
The templates
There were three families of generated pages:
- Room × style. Seven room types times fourteen furniture styles: "Scandinavian living room", "industrial bedroom" and so on. That's 98 pages, and with nine languages, 882 URLs.
- Cities. 74 city pages. One I wrote by hand. The rest came from a seed list with real market data for each city: average price, median days on market, neighbourhoods, a seller checklist.
- Alternatives. One page per competitor, thirteen in all.
On top of that sat the free tools: a flyer maker, a listing description generator, a photo checker and more, each with its own page. Twenty-plus in total.
What Google did with them
Search Console was not kind. At one point about 913 URLs were "crawled, currently not indexed". Here's where three months of clicks actually came from:
| Page type | Clicks |
|---|---|
| 13 tool pages | 108 |
| Home page | 36 |
| 18 city pages | 1 |
| 16 alternatives pages | 0 |
| Blog | 2 |
| Room × style matrix | 0 (39 impressions) |
The tool pages earned more than everything else put together. The room × style matrix, the biggest family by far, earned nothing.
The reason was obvious once I measured it: any two room × style pages were 93.6% identical. Swapping "bedroom" for "living room" and "Scandinavian" for "industrial" isn't a new page, and Google treated it that way.
The cleanup
So I did the opposite of what programmatic SEO usually means, and shrank the site:
- English only in the sitemap. The other eight languages still exist, but they're not submitted.
- Room × style pages are
noindex, follow. They still exist for visitors and still pass links. They just stopped asking to be ranked. - Thin one-off pages were taken out of the sitemap.
The sitemap went from over a thousand URLs to under two hundred. Over the following weeks the average position moved from the high 40s to the low 20s. The tool pages now bring in about 57% of clicks. The image enhancer alone had 66 clicks, and the grass-greener tool ranks around position 9.
Why tools win
A tool page does something a template page can't: it answers the search. Someone searching "make grass greener in listing photo" wants to upload a photo, not read about it. The tool also lets them try before signing up, so it brings in users, not just visitors. Free tools accounted for 52 of 230 sign-ups.
Not every tool works, though. The flyer maker had 531 views and one download, because downloading needs an account. A tool that asks for something before giving something back is just a landing page with extra steps.
What I'd do differently
Start with tools, not templates. Build one page that does something useful, see if it ranks, then build the next one. Generate pages only where each one carries data the others don't, like the city pages do. And measure how alike your templates are before you publish 882 of them.