Making a Case for the Ignore Parameter
July 5th, 2006 by Michael Gray in Google, Ideas, MSN, Random Thoughts, SEO, Tools, YahooIf you're new here, you may want to subscribe to my RSS feed. Read my top posts or learn more about Michael Gray. Want more frequent updates follow me on Twitter. Thanks for visiting!
I know you search engine folks got together a while ago and agreed on a standard to help fight blog spam. Problem is it didn’t work and it only satisfied whiney cat bloggers. Now if you had involved some of the people who were in the industry things might have turned out differently (see classic old school Nick W Threadwatch post The solution to blog spam) you might have had a little more success. I know some of you googlers read my blog, and because I’ve got a head full of crazy genius ideas I’m going to make the case for the adoption of a new parameter.
The way search engines currently work the three URL’s below are usually considered distincltly different:
example.com/foo/
example.com/foo/?loc=top
example.com/foo/?loc=top&a=1
In actual practice I know if Google sees all three of them have substantially similar content they choose 1 and it gets a snippet and if you’re lucky the other(s) may get indexed just a s single line with no snippet. It works that way if you’re on a nice “high power” or “well crawled site”, if you’re a little guy … well I’d recommend not creating something that might be a problem down the road. I purchased my new log analysis software a few days ago and have been playing with it when I had a “light bulb moment” (you know you get a great idea and a cartoon light bulb goes off over your head). What if we had the ability to mark things so site owners could tell what’s being clicked without it throwing search engines a curve ball?
For example many sites have a navigation banner under the header with links to the main sections of the site. They also feature links to the same pages on the side navigation. Now as a site owner/publisher I want to know are people using the links up top or am I giving up valuable screen real estate? Are people clicking the graphical images or the text under the images? Does the buy button at the top or the bottom get clicked more? Now someone is going to say why don’t you use javascript, CGI or some other script based click tracking method. Well if you do it destroys the value of on site anchor text (with Google’s over reliance on authority/trust sites can rank for all sorts of stuff with good internal anchor text and linking structures). So if I could tag the top navigation bar links with a parameter that lets me know it’s the top, and the side with the something that tells me it’s on the side, everything would be groovy. As site owners, publishers, and designers gain more information about how users actually use their sites they can make improvements and make the web a better place.
Still with me … good … OK now the naysayers will say C’mon Gray you can’t be serious, I mean people use all sorts of stupid-crazy parameter naming schemes how are you going to find something that won’t bork the web? Well because I believe simple things work best how about the ignoreparameter? For example:
example.com/foo/?ignoreparameter=topnav
example.com/foo/?ignoreparameter=sidenav
example.com/foo/?ignoreparameter=logoclick
example.com/foo/?ignoreparameter=topbutton
example.com/foo/?ignoreparameter=image
Search engines will see the “ignore parameter” ignore it and remove it from the URL. What’s in it for them, eventually a few years down the road we’ll start having websites that work better and are designed to match the way people use the web, not the way we think they do or should. Some of you are saying c’mon that’s a big job do you know how much time/money/effort that will take? To that I shamelessly steal a quote from one of those inspirational posters you see in the mall …
You’ll never make a shot that you don’t take …
Sphere It
July 5th, 2006 at 6:24 am
Actually, with Google you can already do this by using a query parameter with name “id”. So if you don’t need “id” for something else, just use it. See Google’s webmaster guidelines, it’s there at the bottom of the technical guidelines section.
July 5th, 2006 at 9:10 am
I’m not sure that the ‘id’ parameter is the best way to do it - you’re likely to get the target page excluded instead
This can already be done - look at Google Analytics, it has an overlay map which shows what links were clicked on how often.
I haven’t checked their js, but I expect they are pulling the mopuse coords of each click and defining where on the page it occurred.
July 5th, 2006 at 11:07 am
“This can already be done - look at Google Analytics, it has an overlay map which shows what links were clicked on how often.
I haven’t checked their js, but I expect they are pulling the mopuse coords of each click and defining where on the page it occurred.”
Google Analytics doesn’t appear to separate clicks when links are duplicated though. For example, when you’ve got navlinks on top and the same navlinks on the side, Google Analytics will display the same click stats for both links…at least that’s how the reports show up for my site.
July 6th, 2006 at 6:40 am
Yeah, serves me right for answering questions at 5 in the morning. Forgot about the indexing part, not good at all….
July 16th, 2006 at 5:50 am
Hmm, you’re right.
OK, htaccess allows for it.
Put link
on your page
and
RewriteCond %{QUERY_STRING} ^ignoreparameter=1
RewriteRule ^ foo/ http://example.com/foo/? [R=301,L]
in your htaccess and you get a 301 that *should* have the bots moving on to the real page address for PR, but your logs will include the 301 for Real Clicks.
Of course, you’ll need custom software to analyse it
(technical point: the final ? just before the R=301 is *important*!! Don’t remove it!!)