You're welcome Gretchen - would love to hear how it turns out in the end...

Are your product descriptions well written?  Do they accommodate "cut and paste" use within Meta Description fields?  Remember there's a character limit to Meta Descriptions - too short and searchers may not be enticed to click.  Too long and you leave it to Google to decide where to cut off...

Otherwise, the concept is a sound one, since the idea is you've said "product copy is SEO rich".

Then again what does THAT mean?  A description that is too "spammy" looking may be a deterrent to a click.

Best course of action is to have human review on that whole policy.

And if you want to automate as much as possible, have the first portion of the product description the exact content you want in the meta description field.  Standardize it. Make it a policy that writers need to keep that concept in mind.

Without performing your own tests, there's no 100% best answer for each specific situation, market or site. And you'll find people even here in the Moz community who prefer remaining in the same browser window and just as many who don't.

So... all I can offer is my own experience in UX work - I've found users have an expectation that when they're clicking on a link internal to a site that they remain within the same browser window, but that when they're clicking on any link out (to another web site or social site), that opens a new window.

This is especially important however, when the destination they're going to off-site breaks browser native "back" buttons, where even if you want to go back to the site you came from, you can't.

John,

Thanks for adding all of these great suggestions - I don't do international that often so the full list of methods isn't always in my conscious awareness!

Theory: Google is pretty good at figuring things out.

Reality: Google's algorithms, that go through hundreds of changes, tweaks and modifications every year are a soupy mess and their ability to "figure things out" was proven so flawed last year that along with Microsoft and Yahoo, they came up with Schema.org just to address PART of that reality.

Recommendation: Never do anything that could possibly confuse Google if you don't absolutely have to.

Always place schema markup directly in the position on the page where you want the content to appear if it's content specific - wrapping it around that content. So if your business name and address are in the main content area, that's where you place the schema code.  It's literally a wrapper just like a CSS div would be, or an old-school HTML table, but not for display purposes on-site.

EDITED 11/14/2013 based on a question from Oliver (below) regarding situations where markup is located in the "head" area of the page:

Exceptions to "in-body" markup:

As is the case with any structured markup solution, there will, from time to time, be cases where certain, specific elements go in the "head" section of the code.  Anything that applies to an individual page in its entirety, and does not limit itself to an element of content within the page does, in fact, belong in the "head" area of the page code.

Its absolutely possible that's what's happening.  You cannot rely on Google's system being barred from crawling anything on your site, no matter how well you code it.  Even if you blocked the URL with nofollow, it would not stop the bot.

Another factor is if all your content is in English (as your URL structure suggests it is). Google does a terrible job of discerning separation of international content when all the content is in the same language, on the same root domain.

Proper separation in a way Google can't confuse is vital.  Since I expect you do not intend to change the language across sites, your best action would be to migrate international content to a completely different domain.  At the very least you can then use GWT to inform Google that "this domain is for this country", however if you want to be even better off, you'd host that other content on a server in that country.

Great new Rhys!

Google is less clear most likely because they're still mostly stuck in the belief that they shouldn't reveal clarity and expect site owners to figure it out.  Which inevitably leads, every year, to more and more "what used to be acceptable isn't" complaints.

Except some tactics never were acceptable and Google's just now getting around to addressing some that they previously never considered or never got a chance to.

My latest effort is all about "does this look natural".  That of course, is then filtered through "does this look natural as Google views things in their algorithmic attempt to emulate a human's perspective.

yeah it's annoying that Yelp specifically states it's against their TOS to actively solicit reviews this way, yet they are perfectly happy if you display their "We Yelp" stickers all over the place.

And oddly, places like the BBB's stand-alone "Trust-Link" reviews site is the exact opposite. They encourage business owners to seek reviews.

The danger of reciprocal reviews is being flagged because they're potentially unnatural.  So if there's a one to one parity (every business that reviews another business gets a review by every one of those businesses), that's a serious concern to avoid happening. Same goes for reciprocal reviews that are always the same (4 stars each way, for example).  Too easily spotted as suspicious.

Also, if there's a concerted effort and "conspiracy" to get reviews generated (a bunch of companies join a pool of companies to "agree to review each other"), that could lead to unnatural results.  So it's a very cautious process to even consider.

The other issue is - if a business that participates only or mostly only reviews other businesses in that group, that's highly suspect.  Reviews should be spread out across a wide swath of other businesses NOT in the group, and every participant would need to have their own set of reviews to other outside businesses so no unnatural pattern emerges.

Other than that, it's perfectly valid to review other businesses when you've genuinely done business with them.

It's also perfectly valid for an official business account to review other businesses, since they're business to business transactions.  And thus, no need to have a separate account just for the sake of reviews.  (All reviews should be from an account that has a holistic profile regarding the activity on the account.  It shouldn't be mostly, or all reviews and no other activity).

While it's true that in the overwhelming majority of situations, sub folders are the best solution, I'm going to say that purely from the very limited information shared so far, having sub-domains is far better than having full-blown individual domains, and though not necessarily as good as sub-folders, its still better than the current domain model you have.

It needs to be executed REALLY WELL - with extremely careful thought and consideration given to navigation and cross-domain linking.  However, simply by having subdomains, you instantly let every prospective visitor understand they're all part of the same root domain. That alone boosts your trustworthiness in a BIG way.  And Google does a fair job now at understanding (and in turn providing SOME ranking value) to the root domain from subdomains.

Just don't link to every other subdomain from every other one.  Because that will instantly KILL your SEO.

JavaScript is one of several technologies that offers severe limitations in search engines and their ability to properly see content, then just as important but often overlooked, properly and cleanly evaluate that content from an SEO perspective.

Specific considerations:

• Google does a "fair" job at discovering content passed through JavaScript (either on-page or at the code level)
• A "fair" job means it's hit and miss as to whether their system can actually find that content
• Whatever content the Google system CAN find via JavaScript is NOT necessarily able to be used to properly evaluate content intent, focus or relationship to other content

So - the best practices recommendation is if you want/need content to be found and properly evaluated by Google (or Bing) do NOT pass it through JavaScript.

And also, if you want to HIDE content from Google, don't assume you can successfully do so via JavaScript either.

As for PHP, its the most widely adopted and utilized web programming language out there.  The language by itself is essentially SEO neutral.  It's all in how a programmer utilized PHP that matters.  In the hands of a programmer that either truly understands SEO or collaborates closely with an SEO expert (who also understands the limitations/pitfalls that can arise with "bad" (SEO-unfriendly) PHP coding, it's a great language.