I'm in the process of consolidating several sites into one. Content that previously lived at http://example.co.uk/ will now live at http://example.com/uk/, content from http://example.fr/ will move to http://example.com/fr/, etc.

There are many technical reasons for making this change, but I'm coming up against opposition from local marketing teams who claim that having .co.uk or .fr (and consequently a national identity) in the URL has powerful emotive value for users, and that moving away from it will hurt adoption and conversion.

So I have four questions:

  • Has anyone ever had manage this type of transition?

  • Is there some reference data you found (or generated) that helped in your decision?

  • Can anyone suggest a reasonably practicable way that I might test the effect of a CCTLD on customer behaviour (without necessarily doing a site-wide A/B test)?

  • Or, can anyone recommend a reputable forum that's more focused on conversion rate and influencing factors?

  • CCTLD? Why do you make me look up acronyms when you have ample space available to introduce it properly? As in "country code top-level domain (ccTLD)" Commented Sep 24, 2012 at 18:01
  • 2
    CCTLD is a relatively common term for people who work in and around the web industry - it's not exactly a secret code. In this case I obviously made the incorrect assumption that my audience would understand what it meant. Also, I didn't want to type it all out. Call it 'embracing the convenience of abbreviation'. On the bright side, you're now a tiny bit more knowledgeable then you were before you read my question, though you probably spent as much time complaining to me about having to look stuff up as you did googling 'cctld'.
    – dennislees
    Commented Sep 24, 2012 at 21:27
  • Doesn't the fact an experienced developer had to ask you what a CCTLD was prove that it's not a common term? Commented Sep 24, 2012 at 22:58
  • 1
    What exactly are we talking about here? Good grief. If we're going to be painfully pedantic, then you really should have referred to @MarjanVenema as as "experienced software developer", which gives her a pass in this case. I'd argue that an experienced web developer should know what a CCTLD is. If I was writing some UI language for a broad user base I'd be more careful about using such abbreviations, but in a question to ux.stackexchange? Please. I know this forum is tough on question quality, but come on...
    – dennislees
    Commented Sep 24, 2012 at 23:20
  • 1
    I honestly can not believe this interaction is happing. I'm going to keep this last comment short as I have a life to get back to.
    – dennislees
    Commented Sep 25, 2012 at 13:18

5 Answers 5


I suggest running limited Google AdWords campaigns in the key markets testing CTR (Click Through Rate) with ads running in equal rotation with the 'Display URLs' of example.fr vs. example.com/fr.

Practicalities and cost of this will depend on your industry and organisation of course but you will need a temporary landing page on example.com, though it can just link through to the local site.

AdWords rules on display URLs here; http://support.google.com/adwordspolicy/bin/answer.py?hl=en&answer=175906

I expect given the small variance the Top Level Domain will drive you will need at least a couple of hundred clicks to see an appreciable difference. If you don't reach a significant result, that's also the answer you need.


Apple.ca cares a whole lot about whether or not they lose a few users. Let's assume that this would annoy 1% of the users, the loss of revenue could be billions each year. Amazon does very acute and fine grained analytics on this kind of stuff, but have a different problem where they need to have separate business entities for other countries (probably hosted somewhere else too).

Personally, I think that your marketing's team reasoning is bogus. If you have a good reason for the change, and they cannot prove their point about not doing it, you're the one with the arguments and facts, all they have are opinions.

  • Arguments not backed by some kind of factual basis are called "a wild goose chase" (with apologies to any geese I might have offended).
    – user67695
    Commented Sep 6, 2017 at 18:00

If you need a concrete example, visit http://apple.ca and see what happens. Spoiler: it redirects to http://apple.com/ca/ .

I admit that this is somewhat vacuous as a proof, but if Apple - arguably one of the most brand-sensitive and image-conscious companies of our time - does it this way, it's because this technique is not losing them any money.

It may be noted that Google does not do this, ( http://google.ca does not redirect), but they have a very different business model. Perhaps the Apple example doesn't apply to your business case.

  • 1
    Thanks. It's difficult when Apple is the example used though, as in most cases an an appropriate response is "Yeah, but that's Apple". Same applies to Microsoft and Nike, who also use the subdirectory approach. When you measure unique visits in millions, you can get away with making a few mistakes when testing this stuff out.
    – dennislees
    Commented Sep 24, 2012 at 21:31
  • @dennislees I agree with your last sentence, but I would argue that this is not a test: they've all been doing it for years, and have continued to do it. Indeed, you will be testing it in your business case, but I think that dismissal argument doesn't really hold much merit as the same can be said for any enterprise that not's exactly yours: yeah but they're {huge/small/in or cater to another demographic/sell another product or service}, etc
    – msanford
    Commented Sep 26, 2012 at 1:38
  • @msandford fair enough
    – dennislees
    Commented Sep 26, 2012 at 23:30
  • So the answer seems to come down to: "it will work out the way it does in your particular case" which kind-of undermines the idea of asking questions. I've been seeing a lot of this sort of thing, especially relative to computer programming education, with 400 High Level Languages and several major approaches. Each one does some things well and others badly. If we can't come to agreement on "good enough" then I don't think this internet thing has much of a future.
    – user67695
    Commented Sep 6, 2017 at 17:59

GoogleWebMasterHelp claims that users feel more at home in local sites and that conversion rates should be higher. He also says it is better for SEO


He doesn't quote his source of info or the numbers but hopefully this will help a bit anyway.


It depends a lot on each country.

In some places, the first thing to try is you country domain, like uk, or my, in some other places first comes the .com and second the country.

In many cases, the people try the .com or the country and if there is no result, or not what they expect, then go to google or the search box in the browser and type the name of the company and follow the link.

If you are doing changes in places where people are used to use the country, then changing it is a bad idea, and yes, it will hurt. If you are doing the changes in places where people use more the .com, then it doesn't matter.

If you have an automatic system to redirect, then it's not too bad since the users still can access the site with their usual bookmark or learned address.

You have to check the server logs for increments on people coming from search engines. Also look for people leaving the home page quickly and probably coming back a few seconds later from a search engine.

If you implement redirections, check those logs to find out if people notice the changes in the url and start typing the new one.

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.