Forum Moderators: open
Should I perhaps move one of the two sites to a totally different domain?
If it is going to be a problem, can anyone suggest another approach which will have the same effect?
With such small changes between the pages, I expect them to be listed individually (not merged as identical pages would be), but for only one to show for any given search unless you click "repeat the search with the omitted results included" or use &filter=0.
So IMO the big question is would the matching spelling be the URL that's listed, and would it rank higher than the other spelling would if it was the only one? I suspect that the answer to both questions is yes.
I would suggest you revamp the .co.uk so it is totally different in design and body copy so Google treat them both as different sites.
e.g. (ignore syntax) if (international()) { this sentence says one thing in one way } else { a different way of saying the same thing }
but not having to write everything twice in different ways, doubling my work.
I could do this in the meta title area, keyword area etc. not just to make the spelling differences but also to make the sites different enough to avoid being penalised. But how far do I need to go to avoid penalisation?
The 2nd purpose of the PHP is that the .uk can include info only relevant to the UK and it is omitted from the .com site. I maintain one local copy of my site and don't have to maintain two variants which requires a lot of management. The differences are handled dynamically via the php.
Remember, no deception or attempt to spam is intended in this approach - it's so people looking for (for example) phrases such as "search engine optimisation services" will find me as well as if they type "search engine optimization services" (note different spelling).
1st: Why don't create a only site with dinamicall content? You should be carefull to keep your PR and you on-page factors for SERPs, but it might work
2nd: Use standard English. Standard English is that variant of English that nobody uses. It exists only to provide you with a standard to solve cases like this. In countries where, at school, English is learned as a foreign language, that variant is the one that is studied. It is a problem for students like me, we learn an English that doesn't really exist, but I think that this is the case where standard could be useful.
I wasn't considering human penalties in that post, just what Google's software would do with those pages.
peewhy:
> You have two sites with very minor differences, Google will drop one and index the other.
An interesting notion. Would this be due to human intervention, or because the URLs would be merged as identical?
Greetings,
Herenvardö
THEN sniff for country by IP block - and return the spelling that suits.
using .com and .co.uk to determine someone's country is certainly not a good solution!
We've created localized versions of our site, in local language, local currency prices, and with a ccTLD domain.
So for a country like Italy, or any with a non-english language, I'm not worried about a dupe content penalty, because the language displayed will be different than our .com site.
But for .co.uk, the only difference will be that prices will be in pounds.
I want to be sure our .com will not disappear from the SERPS in favor of .co.uk!
On the other hand, if you had the time, and there was enough business involved in a seperate .uk page, then doing different content should keep everyone happy, not least of which your bank manager...
Can anyone give me an estimate as to how much content should be different?
vincevincevince (..sniff for country by IP block..) : I never thought of that. Excellent idea. I'd have to figure out how to program it. Know of any example? By the way, I'm not really figuring out which country via .co.uk versus .com, because I make it clear on the site that the .co.uk is dedicated to the UK version and there's a link to click to the international version. Note there is no U.S. version - it's international (outside the UK) and UK.
ITcameleon (..also considering creating a UK version..) : If you want to go down this route (i.e. I'm taking all this advice on board and may reconsider), try using php so each version is created dynamically. I'm keeping my .php secret, but it is not very difficult. Hint : strstr($_SERVER["DOCUMENT_ROOT"]. Write a few simple functions based on that value. Just wrap PHP around the bits that need to be different. Force each page to be different and make sure each pair of meta titles, meta descriptions, meta keyword areas and main body are different. Much better than having to maintain two versions of every file.
alpine (..I've got the same issue right now...) : I also use my .php technique to keep the info that is only of use to my UK audience limited to the .co.uk site.
glengara (..this is where PR comes into the equation...) : Very good point. Don't think it will work 100% because if someone's search involves, say, "optimisation" spelled the UK way and they get heaps of results (which they often do) they probably won't click on the "did you mean" results, especially if they're mainly interested in UK sites anyway.
RobinC (..I'd suggest using the browsers supplied accept language header..) : But I want people to find me on either spellings when they do a search. Being listed for only one spelling variant won't help me. Useful info though!
Vec_One (..how much content should be different?..) : See my reply to ITcameleon. Search engines could save a checksum for each area of your page (checksums are unique per content) and if they spot two matching checksums they'll flag duplicate content, investigate and may penalise you. I'm not saying that they do this (only they know), but I'm guessing that they could do now or in the future.
Thanks for all the advice. I'm pursuing my .PHP technique for now (but I've forced more differences) and it remains to be seen how my pagerank will be affected. Probably a Google Dance this weekend. At some point I might switch to vincevincevince's suggestion if someone could advise me on how to program this?
