Redirects

John recommends using redirects, canonical tags, and consistent internal linking to the primary page to solve duplication. He says Google are against using robots.txt to prevent content duplication, because Google can’t recognise the pages are duplicated if they cannot crawl it.

Google say there is no PageRank loss for 301 or 302 redirects.

If you want to remove a mobile site because you now have a responsive site, you should ideally redirect to manage bookmarks for users, but Google doesn’t really care as they’ll drop the mobile URLs when they recrawl the desktop URLs without the mobile rel tag.

If a page is a redirected, Google won’t see any robots meta tags on the page, although they might see a noindex in the headers.

In order to redirect fragment URLs, which can’t work with normal 301 redirects, you need to use JavaScript redirects, or canonical tags on the pages.

Redirect chains cause latency which can slow down crawling, particularly if there are more than 5 steps which will be rescheduled to be crawled later.

For 6 month temporary URL migrations, a 301 redirect is better than 302 to speed up the move.

URLs which are 302 to the home page on any large scale will probably be treated like a 404 and dropped from the index, and pagerank won’t be passed.

Even if a page has been unavailable for some time, then you can still redirect URLs to pass on the authority.

Whilst describing how 301 and 302s are indexed, John seems to suggest that a redirect chain doesn’t pass less PageRank than a single redirect.