5

Question

Are there techniques to let search engines know a page has moved other than 301 redirects?

Background

Hell Migration

We recently migrated our knowledge base from one vendor to another and it's gone very poorly from an SEO perspective. The pages don't do anything you need for SEO to work, and the new platform has severely limited options to address the situation

The URL Journey

1. Original Article: /en-us/Getting-Started

It started with vendor 1, oh how we loved you and your ability to woo search engines with you well structured content and great search results.

2. New Busted Article /article?id=123

Then the new vednor came, it worked and passed all the user tests, got testing sign-off and headed to production. None of us were wise to the fact the new page was dynamic and slow and crawlers couldn't find titles or descriptions for the page.

3. /en-us/Getting-Started -> 301 /article?id=123

So we made the ominous mistake of 301 redirecting all of our old and good vendor articles to the new awful vendor articles. Along with our sitemap and organic and direct traffic from the old url a fair amount of crawlers made the switch to the shitty version of that page.

4. /article/Getting-Started

Fortunately we were able to figure another approach and the search engines are much happier with their cached snapshots with proper titles and descriptions. We've submitted new site maps and the good pages are showing up.

How can I deal with the Busted URLs

Unfortunately we're still seeing results for /article?id=123 with their useless titles and all of our rank they've stolen from the original good vendor.

Option 1 301 Redirect - Not Possible on Platform

I'd love to use a 301 redirect than all the original ranking finds it's home on the page that gives google decent info to work with.

Option 2 rel canonical - Too Slow

I could add rel canonical to tell searchs the good stuff is at the new url, and merge our ranks. But the page is so slow i don't expect the crawlers to actually see it.

Option 3 stripped page, rel canonical, client side redirect - Maybe?

I could strip down the page to just a rel canonical link to the good page, and then redirect to the canonical. But then the content on my supposedly duplicate pages won't look the same at all, which seems like might cause problems.

What should I do

Option 3 is the only option I can see, but I have zero idea how google feels about the page with the rel canonical link and no other content by a client side redirect going to the new page.

Is this a workable approach? Or should I just give up on this one.

Why no 301 Redirects

For the curious, this is a Salesforce "Lightning" Community, kind of a CMS on steroids with a Salesforce connection, which is is in turn built on top of a product call Force.com sites. There is some basic support for 301 redirects on the Force.com sites, but they aren't supported for routes in the "CMS" part of the platform. So I was able to use 301 redirects when we migrated vendors, since that url pattern was not in the "CMS" path, but I can NOT for the implementation that is having trouble with crawlers.

  • 3
    What does "No Platform" mean? Does that mean that your web site host doesn't give you the ability to use 301 redirects? – Stephen Ostermiller Sep 21 at 11:15
  • Welcome to Webmasters! Despite all the detail, afterall, you did a fantastic job in asking your question, I am still a bit confused. I have to admit. But then again, I am still trying to drink my coffee. [insert grin] It may mean I need to wake up more. Cheers!! – closetnoc Sep 21 at 14:40
  • @StephenOstermiller geez that reads terribly with that type. Should have said "Not Possible on Platform". Thanks for the pointer, added a paragraph explaining why 301 redirects aren't possible. – Ralph Callaway Sep 24 at 17:24
  • @closetnoc yeah, i get it, this issue has been making my head spin, all the detail is really just to convince people that 301 redirects aren't an option (not that it worked, the only answer i got was to use 301 redirects). The core question is, is there anything else I can use instead of a 301 redirect to tell a crawler a page to use a new location? – Ralph Callaway Sep 24 at 17:26
1

If you are not able to use 301 redirects, I would strip the page and put in a client side redirect.

  • The client side redirect will get users where they need to go, opposed to a canonical which is sometimes honored by search engines and never used by users.
  • Search engines do pay attention to client side redirects. Both JavaScript redirects and meta refresh. They aren't as foolproof as 301 redirects, but they usually work.
  • Stripping the page will prevent search engines from indexing the old page. Search engines don't index blank pages.

In this case you will do the best you can of getting search engines and users to the new page. I guess that throwing in the canonical wouldn't hurt, but I wouldn't rely on a canonical without the client side redirect.


You also need to find a better web host. All good web hosts give you the ability to redirect. I might be able to live without redirect facilities, however the paired performance issues you are experiencing are unacceptable. There is no way to have SEO or user experience without a speedy site.

  • that's great to know that the search engines will notice the client side redirect, just a little worried about them thinking the page is done loaded before the client side redirect (which was the issue with the original page). – Ralph Callaway Sep 25 at 19:33
  • if the client side redirect is in place, is it worth doing the rel canonical and meta refresh tags in addition? – Ralph Callaway Sep 25 at 19:34
  • and for what it's worth the platform is salesforce, so while it's awful in this particular domain, it's got a lot going for it in terms of having the existing knowledge base flow into a public site, and a decent declarative site builder. it's just awful at loading the pages and there is no low level control (i.e. headers, meta tags) – Ralph Callaway Sep 25 at 19:35
1

It’s a little unclear exactly what the issue with the 301 redirects is, but if you previously set them up you should be able to do it for the new pages as well.

Since you now have 3 potential URLs for the same page, you need 2 redirects both pointing to the canonical page. In other words:

  • /en-us/Getting-Started -> 301 /article/Getting-Started
  • /article?id=123 -> 301 /article/Getting-Started

Furthermore, you shouldn’t discount using canonical tags entirely. They may be slow to update but they help in many other situations such as when any useless parameters are added to the end of a URL (sometimes happens with links from social media). It’s definitely worth adding them anyway.

  • yep, that would make perfect sense, but the platform is Salesforce, and sense isn't their forte. added a paragraph explaining that a bit. any chance you can delete this? I'm worried i won't get any visibility now that i have an upvoted answer. And well, the question was what are the options other than 301 redirects. – Ralph Callaway Sep 24 at 17:27
  • Presumably the first redirect listed above works fine now. What exactly do you get on the /article?id=123 URL now? Is there a page there at all, or a 404? Is the content identical to /article/Getting-Started? – DisgruntledGoat Sep 25 at 12:00
  • @disgruntedgoat the content is the same for end users the two pages, but for crawlers /article?id=123 doesn't fully load and has NONE of the content, whereas article/Getting-Started has a static cache and loads the full content. plan is to "retire" /article?id=123, and do a client side redirect to article/Getting-Started for end users, but also need to make this friendly for search engines – Ralph Callaway Sep 25 at 19:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.