Alexander Blom › Comments

Well I've been here a while now but it was so bluddy ugly before... - Lindy Asimus

Haha yeah. This new layout has been out a little while by memory. - Alexander Blom

Hmm like about 1 day maybe? lol - Lindy Asimus

That shouldn't, 'coz a HEAD request is very different from a GET - Yuvi

I suppose caching could be playing a role. - Dave Winer

A little too early for this :P Sorry! http://blog.bit.ly/post... It's made up of mostly "direct" traffic. - Brandon Titus

Maybe they are hits from web crawler bots (not HEAD requests) that WordPress is filtering out but Bit.ly isn't - Sam Webb

There are SO many ways that could go wrong. Content caching, duplicate hits from the same users, etc, etc. And aren't WordPress.com's counts rather on the conservative side of things anyway? - Joel Bennett

Brandon - thanks for that link. Presumably, the HEAD requests aren't in there now. - Hutch Carpenter

Sam - you think there's some other technology now querying these bit.ly URLs? Interesting. I wonder if it's bots related. Gonna lelave a comment on their blog post, link to this discussion. - Hutch Carpenter

I would bet it's Bots not sure why WP is not counting them but bit.ly is. - Jeremy Arntz

Differences in statistics are universal. I run Sitemeter, Lijit and Google Analytics on the blog. They never agree, and they all differ from the raw stats from the logs. SiteMeter is always the lowest number, but it gives good detail per visitor. - Louis Gray

Which is so funny. A layman such as myself would expect it to be straightforward - how many times was the content called for display in a browser. But there's a disconnect between "clicks" and "views". Bots, HEAD requests are an example of that. - Hutch Carpenter

I'm noticing quite a difference between bit.ly clicks and my Google Analytics numbers. I really would like to understand this stuff better. - Daniel Johnson, Jr.

Hutch, thanks for sharing this friendfeed post! Strange that this was not mentioned during the timoreilly webinar; maybe he is not aware of this issue... - Jeroen De Miranda

HEAD requests do not count towards bit.ly click counts. What you're seeing are requests for a particular bit.ly link where the requesting client does not follow the redirect to its final destination. As Twitter continues to grow, the number of bots and other folks consuming data is definitely increasing. You can see this in action if you post a bit.ly link to Twitter and watch the real-time clicks come in. Almost immediately you'll get about 10 hits. - Nathan Folkman

Thanks Nathan - perhaps related to the proliferation of various services tracking links on Twitter? - Hutch Carpenter

Why not try another URL shortener like http://tr.im and eliminate one of the variables. If it is still out of wack, it is probably Wordpress. If not, it's the URL shorteners. - Neill Adamson

Neill - wouldn't tr.im have the same bot issue? - Hutch Carpenter

The whole bot detection issue is definitely something we are both aware of and working on. We're also trying to help educate developers as well. For example, if you need to expand a bitly link, you really should be using our API, and not, as many do, requesting the bitly link and then parsing out the HTTP response headers or body. There's no way for us to know whether or not a requested bitly link was actually followed through to its destination - that's the big issue here in my opinion... - Nathan Folkman

Hutch - tr.im stats shows the breakup of bots vs human clicks in its analytics pages. The assumption here is that bot is well behaved and identifies itself as such. - Atul Arora

Thanks Atul. I'm a bit.ly fan, but I'll experiment with tr.im. - Hutch Carpenter

Goog analytics will not count clicks from twitter.com as uniques -- that's one reason goog counts are lower than bit.ly. I dont know how traffic from twitter clients looks to goog, but i've been told that some clicks from clients are under-counted. - Adrian Chan