Is there any sort of downside for a URL being changeable?

Take any stack exchange question for example. The URL looks something like this:


Simply webmasters.stackexchange.com/questions/3079 is the "true" link, as in you can always access that question by it's ID even if the title changes.

But say the question gets renamed by the user after being indexed by Google, so the URL is now webmasters.stackexchange.com/questions/3079/file-extensions-are-yummy

Can anyone provide insight into the tradeoffs here? I like to have the document title (in this case, the question title) after the document's ID for SEO and readability purposes, but I'm wondering what the drawbacks are for when the last part of the URL changes.

Is it worth disallowing users from being able to rename their document? Do I need to set up some tricky 301 redirects?

3 Answers 3


Is there any sort of downside for a URL being changeable?

Depends. Let's look at your question's URL. Right now the URL is:


As you noted, if you change the question's title the URL will change, too, and become:


Now, if Google has the original URL indexed and that original URL is no longer valid, then we have a problem, both at the SEO level and, more fundamentally, anyone who's linked to the question or bookmarked the question will not have a broken link and get a 404.

However, this is not how it works on StackExchange. It seems that StackExchange looks at the question ID and ignores the title part afterward. As you noted, you can visit http://webmasters.stackexchange.com/questions/5376 and it loads your question. You can also visit http://webmasters.stackexchange.com/questions/5376/file-extensions-are-yummy (or http://webmasters.stackexchange.com/questions/5376/whatever-dude) and it works, too!

So if you change your question's title and the "new" URL is


That doesn't mean your old URL won't work - it still will. As well as the new one.

At this point you may be rightfully concerned that there are duplicate URLs. If there are a number of links to your original URL (webmasters.stackexchange.com/questions/5376/seo-and-changeable-urls) and a number to the new one (webmasters.stackexchange.com/questions/5376/file-extensions-are-yummy) you are going to be splitting the page ranking in Google and other search engines. Fortunately, this can be mitigated using the <link rel="canonical" /> tag.

Google provides a good description of <link rel="canonical" /> in their blog entry Specify your canonical. In short, you can use this tag to say to the search engine, "I don't care about the URL you used to reach this page, please treat it as if you were visiting via this other URL."

On StackExchange, the canonical URL is the URL of the original question title (or so I presume). If you do a view/source on this page you'll see the following tag:

<link rel="canonical" href="http://webmasters.stackexchange.com/questions/5376/seo-and-changeable-urls">

Likewise, if you visit http://webmasters.stackexchange.com/questions/5376 or http://webmasters.stackexchange.com/questions/5376/file-extensions-are-yummy or http://webmasters.stackexchange.com/questions/5376/whatever-dude the canonical <link> tag stays the same. I presume its value is fixed, regardless of whether you actually change your question's title.

So, long story short, if you have things setup correctly then there is no downside for a URL being changeable. But if you generate 404s or do not use the canonical tag (or some other form or URL rewriting) then there are downsides.

Happy Programming!

  • Thanks! I'll look into the rel="canonical" thing. I do have one further concern, however. Shouldn't the "canonical" be simply the URL with ONLY the question ID since the title is mutable (like for permalinks)? If your page shows up in Google search would it chop off the title from the end if your canonical is only the question ID? Nov 12, 2010 at 23:56
  • I just tried editing the title for this question. Stack exchange now shows <link rel="canonical" href="http://webmasters.stackexchange.com/questions/5376/seo-and-changeable-but-optional-urls">, which is the current url, not the original url. I guess I'll go along with this. Nov 12, 2010 at 23:57
  • @Lotus Notes: I would think that the canonical URL should contain only the question ID and not the title. (I don't work for StackExchange, just a fan, so I have no say how it's implemented! <g>) My guess is that they want the URL displayed in the search results to show up with the title. There would probably be a slight hit to the PR when the title is edited and the canonical URL changes, but the original URL is still valid and will now point to the new canonical one, so the PR damage would likely be short lived. Nov 13, 2010 at 0:18

If you change the URL you have a new page in Google's, and every other search engines', eyes. This means that page will be starting over with PR and incoming links every time it changes. That's obviously not what you want.

If you're going to allow URLs to change then you need to make sure you do a 301 redirect from the old URL to the new URL. But that really isn't a good solution if the URLs change more then once as you will then have a long chain of redirects which will result in lost PR as well as the potential for the search engines to either encounter an error or stop following the chain after X amount of redirects.

There is no good solution for this as you ultimately lose something, if not everything, when changing URLs. I'd recommend against this unless the change is made quickly after creation and left to one change only. Anything other then that is asking for problems.


If the URL changes Google will see a 404 when they crawl that page again, so if your page titles change often you could potentially be deindexing and reindexing multiple pages of your site on a regular basis, this obviously wouldn't be great from an SEO perspective. Also if your page generates a lot of links then you change the page title all those links will point to a 404 and become virtually worthless from both an SEO perspective and user perspective. If you have it set so the content remains as long as the ID is correct then there's the potential for duplicate content causing issues.

A hybrid method that I've seen that works well if the pages are generated manually is to allow them to create a "slug" this is descriptive and gets used as the URL so the user can change the name as often as they want but the URL remains in tact.

(Note a lot of websites are setup this and do fine with SEO as long as you're aware of potential issues and make sure they don't get out of hand you'll be ok).

  • 1
    The "description" part of the URL is irrelevant. As long as you have the correct question ID you can get to the same question no matter what the description is, or if you leave it blank. So there are no 404 headers, it's just that the same page could be potentially accessed by multiple URL's. Nov 12, 2010 at 19:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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