Sitemaps during a migration - which is the best way of dealing with them?

[McTaggart]

Many SEOs I know simply upload the new sitemap once the new site is launched - some keep the old site's URLs on the new sitemap (for a while) to facilitate the migration - others upload both the old and the new website together, to support the migration. Which is the best way to proceed? Thanks, Luke

[CleverPhD]

I would say make sure that your new sitemap has all the latest URLs.  The reason people say that you should have old URLs in the sitemap is so that Google can quickly crawl the old URLs to find the 301s to the new URLs.

I am not convinced that this helps.  Why?

Google already has all your old URLs in its systems.  You would be shocked how far back Google has data on your site with old URLs.  I have a site that is over 10 years old and I still see URL structures referenced in Google from 7 years ago that have a 301 in place.  Why is this?

Google will assume that, "Well, I know that this URL is a 301 or 404, but I am going to crawl it every once in a while just to make sure the webmaster did not do this by mistake."   You can notice this in Search Console error or link reports when you setup 301s or 404s, they may stay in there for months and even come back once they fall out of the error list.  I had an occurrence where I had some old URLs showing up in the SERPs and various Search Console reports for a site for 2 years following proper 301 setups.  Why was this happening?

This is a large site and we still had some old content still linking to the old URLs.  The solution was to delete the links in that old content and setup a canonical to self on all the pages to help give a definitive directive to Google.  Google then finally replaced the old URLs with the new URLs in the SERPs and in the Search Console reports.  The point here being that previously our site was giving signals (links) that told Google that some of the old URLs were still valid and Google was giving us the benefit of the doubt.

If you want to have the new URLs seen by Google, show them in your sitemap.   Google already has all the old URLs and will check them and find the 301s and fix everything.  I would also recommend the canonical to self on the new pages.  Don't give any signals to Google that your old URLs are still valid by linking to them in any way, especially your sitemap.  I would even go so far as to reach out to any important sites that link to old URLs to ask for an updated link to your site.

As I mentioned above, I do not think there is an "advantage" of getting the new URLs indexed quicker by putting old URLs in the sitemap that 301 to the new URLs.  Just watch your Google Search Console crawl stats.  Once you do a major overhaul, you will see Google really crawl your site like crazy and they will update things pretty quick.   Putting the old URLs in the sitemap is a conflicting signal in that process and has the potential to slow Google down IMHO.

[MoosaHemani]

Mate nailed it completely!

[CleverPhD]

Found this while looking for a answer for another question could not find this the other day- right from the mouth of Google to not include pages that do not exist in XML sitemaps.

http://googlewebmastercentral.blogspot.com/2014/10/best-practices-for-xml-sitemaps-rssatom.html

URLs

URLs in XML sitemaps and RSS/Atom feeds should adhere to the following guidelines:

• Only include URLs that can be fetched by Googlebot. A common mistake is including URLs disallowed by robots.txt — which cannot be fetched by Googlebot, or including URLs of pages that don't exist.

[McTaggart]

Very much appreciated CleverPhD!