- Get the new site into the Google Search Console and Analytics.
- Go into riceball.com and remove 20% of the content, to reduce the total work.
- Export old site, and import into new site. Keep the new site “not indexed”. This probably won’t cause page dupes for a few weeks.
- Prepare some regex 301 redirects.
- Find all the URLs that are rooted at /. These may require specific redirects.
- Create some generics that will override the old ones. For example /pages/(.*) -> /pages/$1
- Deal with the 404s at the new site.
- Test
- Spot test for 301s working.
- Look for top URLs in Search Console and test those to get full coverage for all the good-SERP urls.
- Allow search engines to index the new site.
- For successful URLs, remove the original content at riceball.com.
- Goto 4.
What Really Happened: WordPress Export
July 28, 2019
This port was becoming too much work, so in early July, I just did a huge export from riceball.com and imported to technote.fyi, and then removed the old WordPress content from riceball.com.
The sitemaps changed, instantly. The Google index, however, has been taking a while to change. As of July 28, which is around 2 weeks after the import, Search Console says “48 submitted and indexed” and “520 indexed, not submitted in sitemaps”.
Most of the 48 “submitted” are from very old pages, but several pages are still from the old sitemap that should have dropped out.
The 520 indexed fell out of the sitemap, but got reclassified by Google as “not submitted in sitemaps”, despite the fact there was a 301 redirect.
This last thing is weird. The last indexed page I see is on July 26, but I know that it was redirecting to the same article on technote.fyi site on that date. So Google reclassified the URL, but didn’t accept that the page moved to the new domain.
Perhaps these are inbound links?
Nope. Only 16 pages are still linked, with a total of 1,296 links.
The 7 day performance report on July 28 are
riceball.com: 1.61k impressions, 48 clicks.
technote.fyi: 500 impressions, 8 clicks.
So riceball.com still has the ranking advantage.
The Index Seems to Hold Onto URLs
I’m looking at the search results for site:riceball.com, and then clicking on the “cached” feature of each article so I can see the cached version of a page. On the first page of results, several of the pages are copied from technote.fyi, not riceball.com.
So Google caches the latest page, but the entry in the index still has the old URL.
That’s interesting: the system has URLs separate from page content and the actual URL.
As I dug further down the SERPs, the caches showed older page, with different WordPress themes.