Long tail SEO and Google

[plogger]

I have a client that is considering an SEO move that I'm not totally sure about and wanted to see what others think before I make a recommendation. Any feedback, particularly relating to actual experience, is greatly appreciated in advance.

SEO Objective:
Improve long tail keyword rankings with Google.

Background Info:
Large site (several million page views/mo); majority of site made up of categories (think widgets) with hundreds or thousands of paginated listing descriptions under each category or sub-category; each listing description links to a unique page describing that widget in full detail; these category pages with associated listing descriptions as content already rank well for 50+ head terms; approximately 10,000 targeted long tail keywords have been identified; these are primarily derivatives of the head terms; a page has been created for each long tail keyword targeting that phrase; the long tail targeted pages have basically the same content as the main category and sub-category pages, with the exception of unique page titles, meta tags, and headers; client now wants to add links to each of these long tail pages in their XML sitemap.

Example:
You have a category page (www.example.com/category1/) with several hundred ordered listings as content (paginated of course). This category page targets 1 head term. Add 50 additional pages that target related long tail phrases (www.example.com/category1/key1-key2-key3/). Each additional page has the same content as the main category page but with unique page title, meta tags and H1 header. Add links to each of these 50 new pages in the XML sitemap to hopefully get picked up by Google.

My Concerns:
Duplicate content issues; possibly perceived as black hat; keyword cannibalization.

Questions:
Could this really work and improve their long tail SERP rankings?
Could this backfire in a major way?
If yes, to potentially backfiring, do you think this is even worth experimenting with, say trying it out on a few dozen pages to see what happens?
Other suggestions for a dynamic SEO approach to 10k long tail keywords?

[Robert Charlton]

Hi plogger, and welcome to WebmasterWorld.

I'm always amazed at the kind of schemes that some clients come up with.

This sounds like an awful lot of duplicate content to me, and yes, I agree with your concerns that it could backfire. In particular, the scale and intention of the approach would concern me. Google is big on intent, and this is a scheme that's easily spotted.

While Google will normally just filter out duplicate content, the amount of dupe material here under many changed titles might raise a flag and possibly a manual inspection. This is also such a waste of resources that Google might not crawl your site very deeply or often.

Also, I don't think it will work. The long tail content on the page needs to reflect the title. If it does, and you're ranking for your core terms, then inbound linking and Google's understanding of word relatedness should take care of the long tail. If it doesn't, and you're trying a multitude of titles to compensate for that, Google won't rank the pages. This does not mean, btw, that you should stuff keyword variants into your content.

I think it's helpful here to think of this in terms of user experience. If a user came across these paginated pages and started browsing them, it would be a waste of the user's time... perhaps so bad that it would leave a bad taste. Similarly, a page stuffed with keyword variants wouldn't be a good reading experience.

I can't think of any reason why Google would like the setup. This is the kind of "SEO" that inspired Mayday.

[plogger]

Thank you Robert. You basically corroborated what I was thinking, and I appreciate it.

In terms of user experience with the new pages, it's actually not as bad as you might think and still quite relevant to what the user searched for. That said however, I don't think it offsets the potential negatives of duplicate content and, more importantly, the obvious attempt to manipulate Google, which really cannot be denied.

I'm going to recommend they abandon this plan before it gets them into trouble, and we have to spend countless hours trying to "fix" what they broke. I just hope they listen.

[Robert Charlton]

In terms of user experience with the new pages, it's actually not as bad as you might think and still quite relevant to what the user searched for.

I'm not talking about the experience of coming into the page via search. From what you'd said, I'm sure that would be OK. What I'm talking about is what happens when the user starts browsing the site and comes across many pages which are too similar to be interesting.

I'd commented on this specific point, along with several others related to longtail, in this discussion about titles from 7+ years ago....

Title Tags: A badly written title will sink your site
How to sabotage your web site without even knowing it
http://www.webmasterworld.com/forum34/429.htm [webmasterworld.com]

Also, on some types of sites, it's hard to just keep adding pages. If you're building pages only to target phrases, surfing your site can turn into a very unsatisfactory user experience. I feel that it's not possible to build an interesting site composed of pages targeting lots of variations of phrases without substantial interesting content to motivate them.

[freejung]

I don't have personal experience with this, having never tried it, but I think obviously Robert has a very good point.

You might try an alternate approach though, that might be more labor-intensive but may pay off big in the long term.

Consider this: why do you have so many widgets in each subcategory? Are all of the widgets in each subcategory pretty much identical? If so, why bother listing them all? If not, would it be possible to actually split the subcategories up into meaningful sub-subcategories? Could those sub-subcategories corrsepond to your long tail keyword variations?

Another way to express this would be to ask if there is a reason why searchers would use one particular long tail variant instead of another. Are they actually looking for a particular sort of widget, that is distinct from the other widgets in that subcategory?

If so, it might be useful to users to subdivide the subcategories into genuine meaningful sub-subcategories, with unique listings for each sub-subcategory. This creates a lot of pages with unique content that could be targeted at the long-tail vairants, and also would create a better user experience because the user doesn't have to dig through hundreds of pages of blue widgets to find their blue fuzzy polka-dotted widget.

------------
Edited because I realized plogger indicated that there were already subcategories, so I had to change my terminology to sub-subcategories.