Error reports showing pages that don't exist on website

[SOM24]

I have a website that is showing lots of errors (pages that cannot be found) in google webmaster tools.  I went through the errors and re-directed the pages I could.  There are a bunch of remaining pages that are not really pages this is why they are showing errors.  What's strange is some of the URL's are showing feeds which these were never created.  I went into Google webmaster tools and looked at the remove URL tool.  I am using this but I am confused if I need to be selecting "remove page from search results and cache" option or should I be selecting this other option "remove directory"  I am confused on the directory.  I don't want to accidentally delete core pages of the site from the search engines.

Can anybody shed some light on this or recommend which I should be selecting?

Thank you

Wendy

[Millermore]

For me personally, on Wordpress I use the Yoast SEO tool and I went through the tutorial on the Yoast website. He shows you how to eliminate a lot of the duplicate content that automatically gets created with all Wordpress websites. Once you noindex and get rid of all the unnecessary archives and all that, at that point I would recommend going back to the error report and see the difference and see if those pages keep coming up. If they do, just simply 301 redirect them to another page on your website. Then check again after you redirect them and see what you're left with. Sometimes it takes a couple weeks to reflect from what I've seen. Not sure if this is the exact issue you're having, or if you're even using Wordpress at all, but it sounds like if you are this might help you as it helped me get my errors down to zero.

[SOM24]

Thank you for your reply.  You are correct the site is in wordpress.  The long story of this whole situation is this....I had initially built the client a wordpress site, things were fine with, traffic and business was good for him.  Then one day one of his employees suggested that her father build him a new site that was more graphically pleasing (rather than saying can you please update the graphics on the current site) so the father built an entirely new site on Joomla (I didn't find this out until he was launching this new Joomla site)  The guy also went and changed the domain to a www. the original site I had built had no www on it.  Fast Forward.....I have re-built the site in wordpress, went back to the non www version.  The errors I am getting I have 301 re-directed where I could.  I have also in webmaster tools changed the site settings to the preferred domain.(the new site) I fetched all the new pages in Google.  I have submitted new site maps.  I'm down to 82 errors.  The errors are showing pages that do not exist and to re-direct those pages I don't have pages that would really make sense to re-direct them to.  I'm wondering now if I'm to a point where I need to "remove the URL's" as offered in Google webmaster tools...???  What do you think on this?

As for the rel-canonical....I understand why I would use these.  I see in the Yoast plug in where I can insert the rel-canonical.  My question is this:Do I insert the rel-canonical on the page that is correct?  So I go to my correct website (the non www one) and go to that page that is correct and I prefer the engines index that page and insert the rel-canonical on that page that is the preferred one?  Or am I to go to the non preferred page and insert the rel-canonical on that page so when search engines see that wrong page they see the rel-canonical tag showing them the correct page to index?  I looked at a video by Matt Cutts and I wasn't clear on which page I put the rel-canonical on (old site or new one?)  I don't have access to the old site so this is why I was thinking maybe just "remove the URL's" as offered up in webmaster tools.

Your input?  I really appreciate your help.  Thank you

[Millermore]

Rel=canonical is used more when you have duplicate content. If you have the same post or page in two areas, you can use the rel=canonical tag to tell Google where the original of the duplicate is. It sounds like you don't need rel=canonical in this situation.

It sounds like you have 80-something 404 Page Not Found errors. I would use the "Redirection" plugin with Wordpress. Take each URL that is giving you the 404 error in your report, and redirect each one to the most relevant page associated with what was supposed to be on the page that is giving the 404 error. If there really is no relevant page at all, I would just redirect it to the homepage. In my opinion, it's better to have it redirect to the homepage than to have the user land on a 404 page. I would do that for every 404 error you are getting. Doing this, I don't think you'll need rel=canonical at all.

[SOM24]

Ok this sounds good because it's not exactly duplicate content so it would be better I agree to do the redirect.  I downloaded a redirect plugin yesterday that worked pretty good.  I noticed that some pages have already redirected people (the tool has an area to see this)

Final question for you, what are your thoughts on that "remove URL" option in Google Webmaster tools?  Wasn't sure if that would be better than a 301 redirect for these remaining senseless errors.

Just curious on your thoughts.

Thank you

[Millermore]

I would avoid using the "remove URL" option in GWT. The 301s are more ideal in my opinion because let's say I have that old URL posted on my website somewhere, and now it's going to a 404 page. When you redirect it, people will be taken to a different page, and you don't have to worry about having me update the old URL on my website. The link will work, it will take you to an active page and can get you some traffic. However, the "remove URL" option won't give you this same benefit.

Here's a helpful link straight from the source on when NOT to use the Remove URL option: https://support.google.com/webmasters/answer/1269119?hl=en