Is that normal for a staging subdomain and it will lead to proper 404 link when a real domain is linked to the project?
I’m asking this, because a SEO company gave advice to be fixed on the staging domain…
No that’s how it would behave in production, and it’s a legit approach.
Querystrings are a weird corner of SEO space because sometimes they describe a unique page, like ?article=23 might, and sometimes they are something else like gclid tracking data, filter params or something else.
In Webflow, the origin+path pretty much always describes the SEO-able page. Anything on the querystring is more like filter content- including those collection list *_page keys, which you can have up to 20 of on a single page.
Finsweet Filter has a similar querystring ability e.g. ?category=fruit. If I change that to ?category=doesnotexist I wouldn’t expect a 404, I’d expect to get no results.
I don’t think Webflow or Finsweet’s choice to use those querystrings as filter params is a concern for SEO;
The canonical on the page doesn’t use it, if you’re using auto-generated canonicals
The sitemap doesn’t contain them, if you’re using auto-generated sitemaps
There are no links to non-existing page numbers on the site, unless you’ve manually added them
If you have two 10-page collection lists on the page, and query ?(list1)_page=5&(list2)_page=20 you’d still want to get your valid page results back, with list1 showing page 5 and list2 showing no results.
Personally I avoid any efforts to SEO paginated collection list, and instead invest that time in the collection pages they link to which are far more numerous and keyword-relevant.
In most sites I use CMS Load to tidy up the UX for paginated collection lists, and by default it simply hides that paginated querystring too, which is elegant.
