Forum Moderators: open
If you are fairly restrictive about what's including (e.g. by having predefined tags as opposed to allowing a free entry) then IMO you're very unlikely to run into problems.
Of course, one disadvantage of tag clouds is that relative importance is displayed visually (i.e. with a larger font size for most frequent tags) which isn't conveyed to search engines.
Of course, keywords used most frequently in tags will also be linked to most frequently from articles, so tags themselves need not suffer from this problem.
Any input on this guys ?
When I used the tag cloud widget on the blog, each of the tags created a new URL which showed those articles associated with that tag. The article also could be found at the permalink URL. Now the article shows up several times on my blog.
Like I said, I've only began playing with the blog software, but I noticed this as soon as I turned the widget on and clicked a link in the tag cloud.
Does anyone have more experience with this? Does this actually qualify as dup content?
If we were having this discussion pre 2007, I may have had some positive input on this. Right now, in 2008 and beyond, my goal would be to prevent all that "junk" from being seen by the SEs. As you've stated, there may be duplication issues to contend with and all sorts of other "trivial" crawl challenges.
Then you get the "relevancy factor" into the mix and things get a bit distorted.
Do Tag Clouds work for users?
They look more like a Keyword Free For All to me.
caveman's characterization of these pages as the unifying hubs of UGC sites makes sense. Has anyone seen tag clouds help in search on more structured sites, UGC or not, where you'd want them in effect to add modifiers to existing categorizations?
Right now, in 2008 and beyond, my goal would be to prevent all that "junk" from being seen by the SEs.
Different for everyone, but I would consider nofollowing the tag links.
I'm wondering whether there's a way of making tags for subcategorization work. If the point of the tags were to create subcategory pages based on user feedback, wouldn't using nofollows prevent the engines from picking up on those pages (which, I agree, may or may not be seen as dupe content)?
Depends of course on implementation, and whether you ultimately hope to get traffic to the subcategory pages, or whether you're happy with traffic just to the tag cloud pages.
I've been involved with this issue on a number of sites lately, from large to small. Generally my POV is that I love tag pages in certain circumstances for user reasons, and hate them for SEO reasons.
I don't believe the engines want to show them. G has indicated in at least one official place that they don't necessarily treat them differently, but they are essentially thin pages with lots of kw links - the kind of page that continues to fall out of favor over time, especially if there are more content rich pages to show from the same site for a given query. Plus we know how the engines feel about indexing and ranking search results pages, which is essentially what these are.
Also, as a searcher, if I notice a result that is apparently a tag page I virtually always skip over it. (Really? That's the best result a SE can offer? Blech.) And I could be wrong but I think I see far fewer tag pages high in the SERP's than what I saw a year or so ago.
Then consider the scenario of smaller to mid-sized sites. If your site had 100-200 pages, would you want users creating hundreds if not thousands of linked and indexed pages for every tom-dick-and-harry kw combo then can think of? It's different for very large sites.
I do believe there are excellent ways to harness the power of tag pages on any kind of site, but each situation is different and as you say, the specifics of the implementation need to be very well thought through IMO. Whether the site is large vs small, info vs asset vs social vs e-commerce, profit model, all needs to be factored in.
[edited by: caveman at 7:45 pm (utc) on Dec. 13, 2008]
tags are nice and flexible, and really work well on a personal "organising my data" level. Conceptually, they should be good for SEs, particularly for semantic associations.
Practically, I stuggle to see how they would avoid being 'spammy'.
IMHO its like meta keywords- great if not abused. And then the real world kicks in...
[edited by: Shaddows at 12:44 am (utc) on Dec. 13, 2008]
thoughts?
