Redirecting my main index page - is this bad?

[james007]

I have a website which has separate front pages for some countries.

If you visit www.example.com it will redirect you to either www.example.com/uk or www.example.com/ie or www.example.com/global.

This is OK, and incidentally is how The Guardian newspaper works. But I was thinking it might be bad for SEO for the front page to always hit a redirect.

Instead: should a visitor come from the UK, I was thinking about www.example.com being a copy of the UK front page, with a CANONICAL of www.example.com/uk - so it returns the correct UK content, with a canonical to the bookmarkable page, but doesn't have a redirect at the start.

Are there any benefits/drawbacks from doing this?

[TheMadScientist]

I think would probably leave example.com/global (or example.com/ with the global page information) as the "home page", then use <link rel=alternate href=url hreflang=code> on the page, so the default landing page would be /global. Then have a "set your home page option" with a cookie-based redirect to the right version for them so next time they visit they go straight to the right place rather than redirecting the home page to different language variations by guessing at someone's geo-location and which version they want to see.

Here are some of the official Google answers if you haven't seen them:

[support.google.com...]
[sites.google.com...]

[james007]

Thanks for this. Yes, that's one way of doing it, too. I don't want to give people the global page by default if I can avoid it - I don't really want to give people a bunch of irrelevant content as their first experience.

I've elected to stick with automatically giving a user the index page that's right for their area; but removing the redirect - so if you are in the UK, you get given the page content of www.example.com/uk when you hit www.example.com (with canonicals and site links pointing to /uk).

You can switch between editions if you don't like, and every index page is linked in the footer.

[TheMadScientist]

I've elected to stick with automatically giving a user the index page that's right for their area; but removing the redirect

I like this idea.

with canonicals and site links pointing to /uk

If I'm understanding what you're saying I'd keep a close eye on the canonicals and see how search engines respond since you're technically geo-targeting on the same URL -- I'm not sure how they'll handle that if the canonical for each version of the page points to a different URL from the root URL.

It might confuse them a bit, because for one visit the canonical for http://example.com/ could be http://example.com/uk and then the next it could be http://example.com/ie, so algorithmically it might look like you can't make up your mind on what the canonical should be, or they might "get it" and have no issue with it.

Note: To some extent what they do with it and how they handle it could depend on how many people do the same thing, because if there are only a few people who do it that way, they probably haven't coded a specific solution for it unless it was something "easily done or necessary" when coding something else to do with canonical handling.

Bottom Line: If you're one of only a few people doing it that way and there hasn't been a reason for Google (or others) to code specific handling for the situation into the algo, you could get some unexpected results from it, so I'd recommend watching for indications they're "not getting it" and adjust as necessary.

[lucy24]

Does example.com/ exist at all, or is everyone redirected all the time? Seems like this would confuse the search engine, because you've then got a headless site.

There was a recent report that the googlebot is occasionally going to crawl from non-US ranges. So far I haven't seen any hard evidence-- such as log entries*-- that it's really doing this. Maybe wiser heads prevailed as they realized that Googlebot UA + non-Google IP = 403.


* That is, log entries quoted here by other people. I've no reason to expect them to try me from other countries; even Google has a finite crawl budget.

[rainborick]

If you serve dynamic content based on the user's IP address, there's a very good chance that the search engines will never see the alternative content because their crawlers generally don't crawl individual sites from multiple countries (even though Google has that capability now). Worse, it seems to me that even if they do ever crawl the alternatives you'll be presenting them with an indexing problem if the variations are all served from the same URLs.

I think if you want to automate the process, you should use redirects to distinct URLs (preferably using a URL structure that allows you to use geo-targeting in Google and Bing's Webmaster Tools. Overall, if you don't address the geo-targeting issue you're unlikely to ever see search engine traffic from other countries. See Multi-regional and multilingual sites [support.google.com] which gives you a better starting point for websites that want to be both multi-lingual and multi-regional.