More of an experiment

but here it goes..

[blend27]

So,

I took over a old e-com site several month ago to rewrite complete code base and by the glorious request of the new owner there needs to a SITEMAP.xml included as a part of a project reqs.

The site itself is not a bid one, maybe 2000 pages, plus small plus, most are product URIs.

The way is at this point is there are 2 pages per product, 1 for main product page and one with large images for the product.

Like this:

/product.url/productID=nn
/product.url/productID=nnn&showImages=true <<<--these type was kind of thin content pages but have tons of links to them due to HQ images being displayed

The new Product URI will include:

/product-name-productID.html

Both of Old URIs will be redirected to one New one via 301.

This site never had a sitemap and I have full control programmatically when and where old URLs are in the code and when they simply become New URIs.

The way I used to do it on other old sites is to wait at least 4-5 time when old URIs get re-crawled and then flip them to new URI structure from within entire site.

Kind of took a while, but now the Question is about SiteMap file:

To start with: New SiteMap file will be generated for the OLD URIs and fed to SEs before we go with new URI structure(and merged pages), or should we?

What do I do with Old URIs within SiteMap once they are (recorded to 301) re-crawled, should I drop and replace refs with in the code and should I leave refs in SiteMap with addition of New URIs within.

Should I add

<link rel="canonical" href="https://www.aexample.com/newURI.html">

to old set of URIs before they get crawled?

More questions to add later, its a start.

[blend27]

It is a catch22 that I had never to deal with, we just pulled a plug(took the Blue Pill before [webmasterworld.com...] :) in a good'ol days)

[robzilla]

Once you have the redirects in place, there's no point in putting the old URLs in the sitemap. They'll be in the index so Google will re-crawl them eventually and find the redirects. Some old and new URLs may appear in the index simultaneously but they'll catch up after a while.

I'd suggest putting rel=canonical on the new pages in case the old GET parameters are tacked onto the end of them somehow (make sure your own redirect does not do this). I use rel=canonical on just about every page, as I think it's an important hint.

I wouldn't bother creating a sitemap with the old URLs before you make the switch.

[phranque]

the advantage of submitting a separate sitemap with the old urls is that you can track their status as a group in GSC (for as long as necessary or useful)

[blend27]

Now that I'm am looking at it;

To Expand: >>

Currently: Product: Old_URI: /product.url/productID=nn has <link rel="canonical" href="https://www.example.com/product.url?productID=nn"> 
Currently: Product: Old_URI_Images: /product.url/productID=nn&showImages=true has <link rel="canonical" href="https://www.example.com/product.url?productID=nn&showImages=true"> 

Experimental was put in place to 301 to accomplish 'Old_URI & Old_URI_Images' redirected to https://www.example.com/newURI.html

There are no other parameters added to/as-in query-string, blocked for 10 years + as far as I can read the comments, not important, simply 404d, no GAnalytics code what so ever.

So the New Site structure, watch this one:

1. Page A has a link to Old Page B(old URI)
2. Old Page B 301d when followed to Page C(New URI) when visited.
2.1 Old Page B gets crawled N times(lets say N)
2.2.1 ..then the link on page A changes to Link to Page C

All together Now!, this is the same content, just styled a bit diff; Ticklish Yes, and who is Neo to tell me otherwise...

Now the questions about them SiteMap and remember I could control MAPS AutoMagicaly: Add/Remove/Leave In Place URIs as per Doctors orders.

Do ..link rel="canonical".. tags on Page A stay, before B gets tickled a few times stay as A or changed to C URI?.
And Page B gets replaced by Page C in Maps Accordingly?

Hope this makes sense?

[blend27]

P.S. Domain is 20 years old, Old URIs are at least 15 years on the market. Many In-bounds.

[blend27]

I guess to re-phrase the questions would be:

Do I create NEW Sitemap with old URIs and new URIs together and the pick old ones out/off once they are crawled?

AND

Should still I link to OLD URIs from within the site till those URIs get re-crawled?

[robzilla]

Whatever you choose to do, the end result is likely to be exactly the same: the new URLs will be indexed and the old ones will disappear once the redirects are detected. So there's no need to make it this complicated. In fact, I'd say that the more complicated you make it, the more likely it is that you'll confuse the bots.

What I would do is simply swap out the old with the new internally and in the sitemaps, once the redirects are active.

[blend27]

Thank You ROBZILLA!