How to efficiently update the sitemap without losing previous indexes?

Question

I have a sitemap uploaded for my web application. The webmasters has indexed 10% of all the URLs in sitemap so far and marked the remaining as excluded(discovered not indexed). I have made some changes in the app that has resulted in some new URLs and some duplicate URL's for the same content. Now I want to update my sitemap to include the latest URL's while preserving the URL's that are already indexed so far, I have following questions:

1. What is the SOP/Best practice to update the sitemap with modified and new URLs?

2. Should I keep the older sitemap and upload another sitemap with the latest changes and keep multiple sitemaps OR should I replace the older one with the newer sitemap?

3. If I keep multiple sitemaps, what happens to those URLs which are in both of the sitemaps and already indexed from the older sitemap? does google automatically reuses the index or not?

4. If I replace the sitemap, what happens to those URLs which are duplicates(different URLs, same content) of any URL from the previous sitemap and was indexed previously? Does google reuse the index OR drop the index and add the new URL to the queue for indexing?

5. How does 301/302 redirect effects ranking of the page?

Answer 1

While I am a huge advocate of this strategy and have been using it for several years now, I think results are being muddied up a bit.

What I think is being implied is that multiple XML sitemaps will drive better indexation and resulting in traffic increases then a single XML sitemap. I think that may be a stretch and that similar results may be experienced from a single XML sitemap.

What I would be more inclined to believe is that the site may have been experiencing low indexation prior to the XML sitemaps. The creation and submission of the sitemaps (any sitemap) enabled pages that may have been challenged before in being found and indexed to now be indexed. Assuming that they were quality pages with good content, I suspect that there has been a large increase in referring keywords driving traffic to the site, and most likely a large increase in "non-brand" and longtail search phrases.

I do however still agree that this is a great technique, mostly due to the additional level of reporting Google provides. While it must still be taken with a grain of salt, this reporting as a lump sum:

sitemap.xml 40,000 of 75,000 URLs indexed

is far less useful than:

maincontent.xml 98 of 100 URLs indexed

support.xml 20,800 of 25,000 URLs indexed

products.xml 5,200 of 50,000 URLs indexed

etc.

The point being, that products.xml indexation ratio would be alarming and telling...not telling us why so few submitted have been "reported" as indexed, but that those URLs seem to be having a hard time being indexed.

Ideally an XML sitemap helps bypass any crawling issues (if those exist, an XML sitemap is not a replacement to fixing though), so if the pages still aren't being indexed, then it's time to investigate and understand why?

blocked via on-page or robots.txt directive lacking content duplication of content pages have canonical link elements to other pages pages are behind authentication, or content is low-value/duplicate without cookie pages are being redirected etc.

As far how they are broken up, a tree structure might make sense, but there may be other ways as well, based on the section of the site or the type of content...the point is to make the division logical and useful.