where is better to show new url on top of file or bottom?

If I had to guess, I would say the top as the file will be parsed top down and presumably added to the queue in that order.

However, the entire file should be parsed. When I upload sitemaps, Google appears to get all the urls.

I dont think it matters. If SE would scan that file, they would get all urls.

I don’t know what they would do if the read failed before it finished. Scrap the attempt entirely? Use what they managed to get? I imagine they would just try again. In any case as the file can have up to 50,000 url in the urlset I don’t think that’s a major concern.

The optional tags are lastmod, changefreq and priority
I think these are more likely to have an effect than the actual position of a set within the file.

Pure “gut feeling” with no basis on fact, but I think having the url listed in a sequence that reflects a typical recursion would be best. i.e. instead of jumping back and forth from folder to folder, depth to depth, put them in the order a crawl would take.

That means putting the new file where it fits into the tree. Not so easy if you’re editing the sitemap by hand I guess, but AFAIK that’s how the auto apps do it.


I ask because meybe it is better to index first last added content then older.

As Mittineague says, just make sensible use of lastmod, changefreq and priority and you should be fine. If you have a database driven site these are pretty easy to generate.

In theory at least, order is irrelevant in XML.

If you really want to go to town, sort your sitemap in order of last updated, but my intuition is that the whole file will be parsed, prioritised based according to Google’s own criteria and whacked into BigTable before being passed over to a whole load of little search spiders running on thousands of separate machines across Google’s network.

I doubt the order of the source file matters much but I could be wrong.

No matter where you place the site link cz if you are creating an rss file search engine robot could easily find the site url while crawling it. So one doesn’t have to think about that where he/she got to place it.