Yeah, the 301ing could be creating an issue. Google Webmaster Tools really isn't good at following through redirect paths. After hearing that, and reading everything that Matt researched, it's probably worth reaching out to your .edu partners and asking them to update their links to the proper URL for your site (including if it's https now!).

Also, forgive me if I just missed your answer, but have you set up an https profile of your site in Google Webmaster Tools? Google absolutely distinguishes between the two - my site upgraded about a year ago, and I watched traffic drop in one and jump in the other. I keep both now to monitor inbound links that are to either. It's worthwhile creating a profile with and without www as well.

There have been a few people at Moz with similar problems with GSC. People always throw a few ideas around: maybe Google is creating URLs to try to find pages that it can't find through crawling links alone? Maybe another site was trying to hack your site by creating URLs they were hoping would trigger certain content on your site (a laughable idea now, but I remember my college professor showing us a site that put cost parameters in the URL during check out)?

However they got there, though, Eric and Chris gave you some good ways to make sure that you're not still in trouble (if you ever were).

Hope this helps!

Whoa, this is a weird one.

I saw that you posted this on Google's forums as well, and they suggested that this might be the Japanese keyword hack. Did you look into that? If that's not it, did you try loading the URLs that are showing up on the Wayback Machine? It's possible that someone who owned this site before your client created these pages.

Either way, the answer is to double check that your 404 pages really are 404ing. If that doesn't remove them from the index fast enough, you can actually create all of those pages, with a noindex tag, add them all to a sitemap, and submit them to Google. But the 404ing is really your long term solution.

Good luck!

Ah, gotcha. Personally, I use Google itself to find out if something is indexable: if it's my own site, I can use Fetch as Google, and the robots.txt tester; if it's another site, you can search for "site:[URL]" to see if Google's indexed it.

I think this tool could be really good if you keep it as an icon and it glows or something if you've accidentally deindexed the page? Then it's helping you proactively.

Hope this helps!

Kristina

Hi Ioannis,

You're in quite a bind here, without a good URL structure! I don't think there's any one perfect option, but I think all of these will work:

• Create a page on your site that links to every article you would like to delete, keeping those articles 404/410ed. Then, use the Fetch as Googlebot tool, and ask Google to crawl the page plus all of its links. This will get Google to quickly crawl all of those pages, see that they're gone, and remove them from their index. Keep in mind that if you just use a 404, Google may keep the page around for a bit to make sure you didn't just mess up. As Eric said, a 410 is more of a sure thing.
• Create an XML sitemap of those deleted articles, and have Google crawl it. Yes, this will create errors in GSC, but errors in GSC mean that they're concerned you've made a mistake, not that they're necessarily penalizing you. Just mark those guys as fixed and take the sitemap down once Google's crawled it.
• 410 these pages, remove all internal links to them (use a tool like Screaming Frog to make sure you didn't miss any links!), and remove them from your sitemap. That'll distance you from that old, crappy content, and Google will slowly realize that it's been removed as it checks in on its old pages. This is probably the least satisfying option, but it's an option that'll get the job done eventually.

Hope this helps! Let us know what you decide to do.

Best,

Kristina

Agree with Oleg above; that's not a good solution if this lasts years, though.

Is there any chance you could serve the old site using the current URLs?

Google uses URLs as unique identifiers, which means that if you redirect to URLs that are only for the old site, it'll think you've created new pages. New pages mean a month or so of low rankings, before your page has a proven click through rate and gets some links.

If you use your current (and future) URLs to load the old site, though, Google won't see that as a site overhaul, it'll see it as content modifications on a pre-existing page.

Hope this helps! If 302s or using the same URL isn't an option, let us know so we can brainstorm more solutions with you.

Best,

Kristina

I would try to keep the same data, but organize it in a way that looks good on a screen the width of a mobile phone. HTML tables render the same way on a phone as they do on a desktop, they're just more likely to get smooshed.

I recommend playing around with your HTML using Google Chrome's Inspect option, and changing your device to a relatively small smartphone. Then build a table that looks good there.

If your table has too many columns, I'd recommend dividing the table into multiple tables, so you have the same data but in fewer columns for each table.

Good luck!

Kristina

For any duplicate content you have between countries, use hreflang to differentiate regions. Google lays out how to do that here.

Hope this helps!

It's for different regions as well. Check out the link I shared. Google lists the reasons for hreflang. The second reason is:

"If your content has small regional variations with similar content, in a single language. For example, you might have English-language content targeted to the US, GB, and Ireland."

I would recommend setting up each country's subdirectory as separate properties in Google Search Console. Then, go to original Search Console, and click on Search Traffic > International Targeting, click the tab Country, and identify which country you're targeting users in.

That should give GSC enough information to not flag the content as duplicate.

Good luck!

What I've usually seen with canonicals is that Google either removes the noncanonical page from its index, or it ignores your canonical and treats them as two separate pages. I haven't seen an example where a canonical lets you get the best of both worlds.

I agree with Nozzle - you can tweak your existing content to target both phrases! Google understands synonyms, so if anything, you're just creating a more all around relevant page.

Good luck!

Kristina