Implementing https in Wordpress

is duplication still a risk?

[samwest]

Last year I switched from a good old html site to all Wordpress. The rationale was to modernize and to employ WP's power in helping me organize a growing site.

I recreated my main pages as pages and kept the blog posts from when we were previously using WP for the blog only.

I installed WP in the root, then redirected all my static pages over to the wordpress pages. All seemed fine.

A few weeks ago, since I already had an active SSL cert, I set my wordpress address and site address to https://

Due to some odd errors with SSL that caused pages to lose their css, I added a plugin called: Wordpress HTTPS which seemed to fix any user level issues.

The problem is that now my traffic is tanking, during a period when I usually do very well (Jan-April).

I did a few checks and found with Majestic SEO that under "Pages" it shows both the http and https of the home page with approx 1,000 backlinks from the http and about 10 from https.

I'm afraid I may again have some canonical issues that Google may be viewing as duplication and penalizing me for this...which is odd because G advises to switch to SSL.

Any advice is welcome.

[Planet13]

I don't know what it is off the top of my head, but I am pretty sure there is a 301 redirect for apache which will change http to https automatically.

also, make sure that the pages aren't being served with both www and non-www prefixes.

BTW: You are going to have to share with us all your secrets for getting so many backlinks.

[not2easy]

There are several things to update when you switch to https: URLs. Remember to go to your WP Settings file and change the URL of your WP URL so that it will only deliver via the new https: URLs. You need to add the site as a new site in GWT and be sure that your canonical redirects send all http: requests to https: via 301 (permanent) redirect.

First suggestion would be to try to access old pages (the legacy content pre-WP pages that were redirected) and check if you need to edit their redirect to include https. Check WP URLs with and without https to see if you can access them. That should tell you where your edits can help.

[samwest]

So far it sounds like I've done all that is suggested above...

I did delete the old WMT account and set up a new https version. Once I saw my search queries on the old http site drop to zero, I discovered that.

As far as redirects...
If I type:

example.com > I get [example.com...]
www.example.com > I get [example.com...]
http://www.example.com > I get [example.com...]
[example.com...] > I get [example.com...]

I also 410'd a few old outdated pages.

Question: Why does Google (in particular) seem to ignore 410 requests? I've had this up 6 months and they keep coming back to the same 410'd pages. This is also true of 404's. You'd think there would be a time frame in which they just give up or just follow the instructions in my site map (which is all https).

A few days ago I turned off 410's and redirection and watch traffic flowing with Wordfence Live Traffic....not seeing as many 404's as before, so many I can nix the redirects and let things evolve naturally from here on out. Good idear or bad?

Thanks for your informative answers.

[edited by: lorax at 12:58 pm (utc) on Feb 2, 2015]

[edited by: engine at 2:13 pm (utc) on Feb 2, 2015]
[edit reason] examplified [/edit]

[not2easy]

There was just a discussion yesterday here: [webmasterworld.com...] about 404s in GWT from pages gone for years, same problem as the 410s. Inbound links that are still out there get crawled so they report it again.

I don't think that is always the case though because I click sometimes to see where the heck a dead, gone, buried and long forgotten page got crawled from and they claim it was found in the sitemap. Well, yes, the 2007 sitemap maybe. Some people see these weirdnesses as evidence that Google is in the midst of deep crawling and new adjustments.

All you can do is mark it "Fixed" (Translation: yes, I know that page returns a 410 - as it should) and wait for the next time.

[Mroffline]

It is a duplicate risk if both http and https versions are crawlable.