Why google stubbornly keeps indexing my http urls instead of the https ones?

[max.favilli]

I moved everything to https in November, but there are plenty of pages which are still indexed by google as http instead of https, and I am wondering why.

Example: http://www.gomme-auto.it/pneumatici/barum correctly redirect permanently to https://www.gomme-auto.it/pneumatici/barum

Nevertheless if you search for pneumatici barum: https://www.google.it/search?q=pneumatici+barum&oq=pneumatici+barum

The third organic result listed is still http.

Since we moved to https google crawler visited that page tens of time, last one two days ago. But doesn't seems to care to update the protocol in google index.

Anyone knows why?

My concern is when I use API like semrush and ahrefs I have to do it twice to try both http and https, for a total of around 65k urls I waste a lot of my quota.

[max.favilli]

Thanks again Dirk! At the end I used xenu link sleuth and I am happy with the result.

[DirkC]

Hi Massimiliano,

In Screaming Frog there is the option: Bulk Export > All inlinks -> this generates the full list of all your internal links with both source & destination. In Excel you just have to put a filter on the "Destination" column - to show only the url's starting with "http://" and you get all the info you need. This will probably not solve the issues with the images. For this the next solution below could be used.

The list can be quite long depending on the total number of url's on your site. An alternative would be to add a custom filter under 'Configuration>Custom' - only including url's that contain "http://www.gomme-auto.it" or "http://blog.gomme-auto.it" in the source, but in your case this wouldn't be very helpful as all the pages on your site contain this url in the javascript part. If you change the url's in the Javascript to https this could be used to find references to non https images.

If you want to do it manually, it's also an option - in the view 'internal' of the crawler you put "http://" in the search field - this shows you the list of all the http:// url's. You have to select the http url's one by one. For each of the url's you can select "Inlinks" at the bottom of the screen & then you see all the url's linking to the http version. This works for both the html & the images.

Hope this helps,

rgds

Dirk

[max.favilli]

Forgot to mention, yes I checked the scheme of the serp results for those pages, is not just google not displaying it, it really still have the http version indexed.

[max.favilli]

Hi DC,

in screaming frog I can see the old http links. Usually are manually inserted links and images in wordpress posts, I am more than eager to edit them, my problem is how to find all the pages containing them, in screaming frog I can see the links, but I don't see the referrer, in which page they are contained. Is there a way to see that in screaming frog, or in some other crawling software?

[DirkC]

Hi,

First of all, are you sure that Google didn't take the migration into account?I just did a quick check on other https sites. Example: when I look for "Google Analytics"  in Google - the first 3 results are all pointing to Google Analytics site, however only for the 3rd result the https is shown, even when all three are in https. So it's possible it is just a display issue rather than a real issue.

Second, I did a quick crawl of your site and I noticed that on some pages you still have links to the http version of your site (they are redirected but it's better to keep your internal links clean - without redirections).

When I checked one of these pages (https://www.gomme-auto.it/pneumatici/pneumatici-cinesi) I noticed that this page has some issues as it seems to load elements which are not in https - possible there are others as well.

example: /pneumatici/pneumatici-cinesi:1395 Mixed Content: The page at 'https://www.gomme-auto.it/pneumatici/pneumatici-cinesi' was loaded over HTTPS, but requested an insecure image 'http://www.gomme-auto.it/i/pneumatici-cinesi.jpg'. This content should also be served over HTTPS.

The page you mention as example: the http version still receives two internal links from https://www.gomme-auto.it/blog/pneumatici-barum-gli-economici-che-assicurano-ottime-prestazioni and https://www.gomme-auto.it/pneumatici/continental with anchor texts 'pneumatici Barmum' & 'Barum'

Guess google reasons, if the owner of the site is not updating his internal links, I'm not going to update my index

On all your pages there is a part of the source which contains calls to the http version - it's inside a script so not sure if it's really important, but you could try to change it to https as well

My advice would be to crawl your site with Screaming Frog, and check where links exist to http versions and update these links to https (or use relative links - which is adviced by Google (https://support.google.com/webmasters/answer/6073543?hl=en see part 'common pitfalls')

rgds

Dirk

[max.favilli]

Mhhh, you are right theoretically could be the crawler budget. But if that is the case I should see that from the log, I should miss crawler visits on that page. Instead the crawler is happily visiting them.

By the way, how would you "force" the crawler to parse these pages?

I am going to check the sitemap now to remove that port number and try to split them. Thanks.

[MonicaOConnor]

Darn it, you are right, we added a new site, not a change of address, sorry about that. Apparently my coffee is no longer effective!

[max.favilli]

As far as I know the change of address for http to https doesn't work, the protocol is not accepted when you do a change of address. And somewhere I read google itself saying when moving to https you should not do a change of address.

But they suggest to add a new site for the https version in GWT, which I did, and in fact the traffic slowly transitioned from the http site to the https site in GWT in the weeks following the move.

[OlegKorneitchouk]

Are you sure? On https://support.google.com/webmasters/answer/6033080?hl=en&ref_topic=6033084 it says: "No need to submit a change of address if you are only moving your site from HTTP to HTTPS."

I dont think you are given the option to select the same domain for change of address in GWT.

[OlegKorneitchouk]

Looks like you are doing everything right (set up 301 redirects, updated all links on the site, updated canonical urls) - just need to force the crawlers to parse those pages more. perhaps crawler is hitting its budget before it gets to recrawl all of your new urls?

You should also update your sitemap as it contains a bunch of links that look like: https://www.gomme-auto.it:443/pneumatici/estivi/pirelli/cinturato-p1-verde/145/65/15/h/72

I recommend creating several sitemaps for different sections of the site and seeing how they are indexed via GWT.

[MonicaOConnor]

Did you do a change of address in Google Webmaster Tools? Http and Https are considered different URLs, and you will have to do a change of address if you switched to a full https site.