NOINDEX,FOLLOW on product pages

[gtrotter666]

Hi

Can I have people's thoughts on something please.  We sell wedding stationery and whilst we can generate lots of good content describing a particular range of stationery we can't relistically differentiate at a product level.  So imagine we have three ranges

Range 1 - A Bird

Range 2 - A Heart

Range 3 - A Flower

Within each of these ranges we would have invitations, menus, place cards, magnets etc.  The ranges vary quite alot so we can write good textual keyword rich descriptions that attract traffic (i.e. one about the bird, one about the heart and one about the flower).  However the individual products within a range just reflect the design for the range as a whole (as all items in a range match).  Therefore we can't just copy the content down to the product level and if we just describe the generic attributes of the products they will alll be very similar.  We have over 1,000 "products" easily so I am conscious of creating too much duplication over the site in case Mr Panda comes to call.

So I was thinking that I "might" NOINDEX, FOLLOW the product pages to avoid this duplication and put lots of effort into making my category pages much better and content rich.  The site would be smaller in the index BUT I do not really expect to generate traffic from the product pages because they are not branded items and any searches looking for particular features of our stationery would be picked up, much more effectively, by the category pages.

Any thoughts on this one?

Gary

[irvingw]

Sounds like it makes sense and that you have thought it out. If the category pages are conversion friendly, sounds like it can be done. But if there is a way you can get the product pages to have unique content I would personally prefer the product pages to rank. By doing what you are suggesting you're putting the point of purchase a click further away, which isn't the end of the world.

[AlanBleiweiss]

One alternate method would be to integrate the products onto their individual "range" pages, with purchase capability and options right there.  You'd need to ensure the overwhelming majority of those pages is still unique, however it would avoid potential confusion that comes from "noindex,follow" being used on a massive scale, which can itself be problematic.  (Google needs to understand WHY there are so many nonidex pages, and what unique links exist on those pages that you want the crawler to follow them for).

[gtrotter666]

Alan Thanks for this. Can I check I understand your comments. Are u suggesting that a large number of noindex, follow pages causes google to lose interest in following the links from those pages? Do u know this to be the case through an empirical study? I like your suggestion of integrating the product purchase onto the category pages. I agree that would be ideal but the products themselves have alot of options and some are designed online so it could end up quite complex. Food for thought though as it would be a good solution SEO wise. I'm just a little concerned on a user ability front. Gary

[AlanBleiweiss]

The issue with a high volume of noindex,follow pages is understanding intent, and potentially having the message be confused.  "We have x pages with noindex, follow - meaning "we don't think these pages are important enough to index, but the links on them are". Except if those links exist elsewhere, on enough indexed pages, what's the point being made?

Is it an attempt to artificially boost the signals for those pages that are linked by saying "look at all these extra links we have pointing to these other pages"?  That's the concern especially since the implementation of over-optimization factors in algorithms. While it may not be Google's intent to devalue a site due to innocent behavior, their ability algorithmically to understand is limited.

Over the past year and a half I've seen more and more situations where Google's many layers of algorithmic decisions have resulted in client sites suffering because of a lack of human review that can determine "this was not an intentional attempt to over-optimize".  I've seen it with internal linking, I've seen it when use of noindex/follow conflicts with canonical signals, and I've seen it where either of those conflicts with robots.txt instructions.

While no single case is guaranteed to be problematic (due to hundreds of factors being evaluated across multiple algorithms), at the same time, as a professional audit consultant I am not comfortable enough to then leave out the consideration where no single case is guaranteed to be safe either.  Thus, my opinion of "best practices" is to "avoid potentially significant problems".

[gtrotter666]

OK, thanks for your detailed response.

I am wondering whether I might just have a dynamically generated URL for the "product level" pages, i.e. page.php?id=1, page.php?id=2 etc and then have a canonical tag that is the same across them all.  I could then limit the product pages to those that are genuinely different in some way.  That way, I can avoid the noindex,follow issue, have very few duplicate product pages and avoid Panda related issues.  Sound sensible?

Gary

[AlanBleiweiss]

Yes that's a very good idea.  It is much stronger a signal in it's execution than the noindex/follow method given my concerns.

[gtrotter666]

Thanks for helping me bounce the ideas around.  Always valuable comments from SeoMozzers!  Have a good day!