Msg#: 4149558 posted 12:13 pm on Jun 9, 2010 (gmt 0)
I'm building a basic form search facility which allows users to select checkboxes (16 total) representing database fields. Ideally I'd like Google to be able to spider the results pages, but I'm realising that in order to create unique title/h1 content (based on the selections) for those pages the tags will potentially end up being quite long and trailing. So I'm wondering where the priority should be on making them unique. How should I deal with this? Would a long trailing title and more concise, but non-unique, h1 be ok, or vice versa? The meta description will also be unique but length won't be an issue there.
Msg#: 4149558 posted 10:02 pm on Jun 9, 2010 (gmt 0)
Just so I'm clear, are you talking about a possibility of 16! (factorial) possible pages - like 2 followed by 13 zeros? Or is there a likely limit to the number of option combinations selected at one time?
That kind of dynamic tag creation can be a nightmare. If you make the pages technically crawlable, then some of them may be crawled - but only crawled, indexed and ranked because of other factors, inlcuding backlinks. Just being a form selection probably wont do it.
As far as the title element, we can see that Google truncates the display around 70 characters or so, but they do seem to index slightly longer tags - how much longer is an open question.
The meta description is not a ranking factor. Rather than create duplicates (and doing it dynamically sounds like that would be inevitable) I'd probably let Google build their own snippet in cases where the page actually ranked.
I'd also be concerned with generating the url for that kind of page. Bottom line, it doesn't sound like the kind of content that Google wants to index - but I say that without knowing critical details, of course.
Msg#: 4149558 posted 11:08 am on Jun 10, 2010 (gmt 0)
Thanks for your reply tedster.
Bottom line, it doesn't sound like the kind of content that Google wants to index
I think you could be right, in the sense that there won't actually be anything unique in the content of the search results, just the way it's grouped according to the search. I suppose I was attempting to create something crawlable oriented around specific keywords.
Yeah there will be a lot of possible combinations, but it's intended as a useful tool for the user more than anything. Should I let Google in or perhaps "no index, follow" might be more appropriate?
Msg#: 4149558 posted 12:48 pm on Jun 10, 2010 (gmt 0)
Sounds like you could be creating a duplicate set of content which whilst it might be useful for your site users isn't really that unique or valuable for search engine users. Probably best to no index, no follow