Welcome to the Q&A Forum

Diane, what I did in a similar situation (though not with anything like as many pages), was to create rules in htaccess for the old to the new url structures using mod rewrite and regular expressions. It's pretty much impossible to say what rules you need without knowing the urls.

That works if your new urls are in some way related to the old ones otherwise you need to do singe redirects for the old urls to the new ones. (Again you can do this all in htaccess, but sh404sef allows you to do redirects too, suggest you ask on their support forums if you don't know how.)

Finally, do you really need to redirect all those old pages? If they didn't have any incoming links, then how much value (i.e. link juice) were they really adding to your site? How much traffic were those pages bringing - if little or none, why not just let Google realise figure out that they're gone and (eventually) drop them from the index?

That's one of the things I was wondering about Nick - how much other (i.e. non-browsing behaviour type) personal information does Google et al gather?

Another type of information is things like  the UK Companies House maintains information of registered companies (including registered office address etc.) - could that be useful for local ranking? (BTW I'm not sure Companies House info is relevant, but there may be other sources that are).

@Dejan SEO, not sure if using offline data would go against Google's 'philosphoy'. I guess there are (at least) two categories of 'offline' information that would be relevant: things that are truly offline (for which an entirely new 'parser' would be required) and things that can be imported and correlated with online data.

I think the postcode example is a good one - if I live on a particularly poor neighbourhood and do a search for 'where can I get a good family holiday to suit my budget' or something like that, a 'better' result for me might be different to someone living in an affluent area...and Google does say it wants to give the 'best' or most appropriate answer doesn't it?

How to turn that idea into something actionable for SEO? Well, I'll leave that to someone cleverer than me!

@digna99 no, please online ID cards!?!

A journalist contact of mine was asking whether we were getting close to the point where 'developers' (for which I read Google) can take account of opinions/interests/actions of people who don't use the Internet (the example I was give was 'poor' people and 'rich' people).

This got me thinking about what (if any) offline signals that Google/Bing etc. use as signals for ranking. I understand that some household demographic data may be used for ad targeting (I think I read that Axciom do that) - but are there any others?

I'm not sure that statement about rel canonical only working within your own domain - if you have some test data/similar that shows this to be the case, I'd love to hear about it.

Matt Cutts specifically says that cross-domain rel canonical is supported, see: the webmaster video on: iwww.youtube.com/watch?v=zI6L2N4A0hA

Hi Alan, posting this question clearly had an effect; within a week the google fairies had sorted themselves out and now only the www version is showing in SERPS.

There's a beta version of sh404sef for joomla 1.6 now (if that's what you're using).

Having said that, if you can move to joomla 1.6 you'll find they've made some effort to make it much more sef than 1.5 - though I did have to do some tweaks to get a canonical tag in there. I'll be using the sh404sef when it's out of beta to avoid fiddling around with menus to create clean urls.

Agree with Damon on the components - if you're using them then check compatibility with whatever SEO extension you use (if you use one that is)

Rand mentioned using cross-site canonicals at the linklove event in London, perhaps some of the SEOMoz staff have some data on that kind of usage that they could share? (particularly with respect to the how closely content needs to match from page to page).

We found 5pm a bit unfriendly to use, ended up with cohuman which is much easier and integrates with gdocs. It isn't as good as 5pm at the high level for project overview. None of the team likes base camp though we have to use it for one of our clients.

will let you know. I'm a bit annoyed since our sitelinks seemed to disappear somewhere along the lines while these shenanigans were going on. hopefully they'll come back when it settles down

I was hoping that might be the case - needed someone with more experience to cast their eyes over it. Thanks!

Hi there, grateful for any ideas on why this is happening:

http://www.google.co.uk/search?q=site:www.vitispr.com

vs

http://www.google.co.uk/search?q=site:vitispr.com

Google seems to be indexing and caching vitispr.com for our home page but the www. versions for everything else. As you can see the second query finds the home page.

Any ideas why that might be?

Other info that might be relevant:

• non-www etc. are all 301'd to www versions.
• moved domains/urls etc. around in March of this year and for a week or we were redirecting to the non-www version
• webmaster tools says 'www' preferred

Thanks!

There was a 'ask Matt'  google webmaster video on YouTube not too long ago where he talked about errors in markup - if I remember right he said it wasn't too important - but of course Ryan's comment is sound - so long as the errors don't break the site for users/crawlers.

Generally we find PR Web is better (and cheaper!), Marketwire has been around longer and I think they may have a stronger contact base in the traditional media.

That's a nice list you've got there, I particularly liked the disclaimer about public relations at the top of the page ;-).

We're putting something together that's similar, but coming at it from a PR agency perspective rather than straight linkbuilding for SEO.

Well, the wedding's over so this may not help the OP directly, but could be useful for others:

• For tech clients we use sourcewire regularly (and find it works very well)
• For cheap, releases that still give half-decent links and get picked up by google news, we often use www.journalism.co.uk (which I didn't see on your list and costs about £30 for the basic service)

As a (slightly) amusing aside, we put an April Fool's blog post up about being appointed the PR agency for the royal wedding - would you believe we actually got a couple of enquiries from journalists asking us for information about it! I dread to think how many more enquiries we would have got had we used a wire.

Thanks Todd, you're very kind. Might be worth marking my response as a 'helpful answer' or the question as 'answered' since it still looks unanswered.

I found somewhere a tip to put some code in the head.php file of your template (sorry can't find it at the mo, but I'm sure some Googling will discover it for you).

I ended up putting this in my head.php  (which I found here: templates/[MY TEMPLATE NAME]/blocks/head.php) BTW I didn't want the trailing slash in my URL - you'd need to tweak this if you do want it.

But you could probably get away with something simpler, there's some discussion about it here: http://www.brucerick.com/joomlas-sh404sef-and-canonical-links/

Oh and feel free to improve my php/regex...I'm no expert