In a regular afternoon check of techmeme.com, I was intrigued to find a linkbait post of a most intriguing kind from Search Engine Land: Get a Free Link from Wired

Sure enough, Wired has a wiki. I quickly hacked up a how to on conducting usability testing, plugging my own Scrutinizer Browser which is an interesting way to empower a novice with expert level ability to observe the thoughts of a test participant.

I didn't pay much attention to what Danny had posted to Wired, but it turns out he spammed them with non-how to promotional content. I did manage to check back, after calling attention to opportunity to my fellow Stomper faculty, Don Crowther. We're in the midst of launching a course on how to do social marketing. But unlike Danny's post, mine served Wired's presumed goal of accumulating a large number of how to articles.

Danny's post has been updated a couple times, so it's a bit late to do a blow by blow. Lots of folks followed suit in spamming wired, consuming Wired server and brand capital for purposes not advantageous to Wired. Editorial is keeping up with the low flow of crud thankfully. Search Engine Land has apologized for inviting spam.

There was an early update that I found quite homourous:

Postscript: Seems like Wired is now calling our test entry spam and deleting it. Plus, Ross Mayfield, Wired's Wiki Editor is incorrectly saying that nofollow doesn't "work" on wikis.

Ross Mayfield, a true technologist, countered that no-follow is inappropriate for wikis. Danny@SEL rebuked this statement strongly and others followed suit. When all the engines jumped on the no-follow bandwagon in January '05, I wrote a post title "Settling for Just Good Enough":

While the recent move across multiple industry players to support rel="no-follow" on links is a positive step, it falls rather short. Vote links, with -1,0, or 1 values, would have been a much more interesting solution to this problem and left room for the community to engender evolution, instead of simply elimenating a threat to the already plotted growth.

I'm confident that blogging tools will soon support this for comments and referrer links, but I regret that the effort will be spent on the most impoverished conception of link typing...

When Ross says that no-follow doesn't work for wikis, he's speaking from a semantic POV, not a SEO/business/untrusted content one. For the wiki to work, a wide range of potentially contrasting viewpoints are possible.

I find it sad that a linkbait stunt forced Wired's hand into going the no-follow route when they might have found an editorial or user feedback route to keep SPAM under control. Just because Wikipedia made what seems a smart decision to no-follow, doesn't mean that's the right solution for Wired How-To -- a smaller and more restrained content space.

I look forward to the day when a major search engine takes on not just a single boolean about the level of endorsement of a link to supporting a range of levels of support, or even relationships.

It's long overdue that analytics packages diagnose the quality of site search and Google announced several additions coming soon to Google Analytics at EMetrics. While GA is late to the game with on site search query term analysis, they are adding some diagnostics for user success with search.

As a partial solution, I historically have set up a funnel mapping search result to clickthrough success by adding custom urchinTracker results to all search results pages and pages referred by search results. Here's what the various views looks like:

Search Clickthrough "Conversion Rate"


Search Funnel Abandonment

The Search Funnel

Search success is a tough thing to measure. Search, then click is a clear success pattern, but what about search, refine, and then click? What if users click several times?

Check out Avi's post for screenshots of the new reports and my post in the webanalytics forum for some of the wider issues.

Additional features are coming for more detailed event logging, including default outbound link tracking, but the details are not yet available.

I'm just back from SIGIR 2007 where I interviewed Jim Jansen on his long term research on query strings as well as more recent work on the nature (quality, perception) of search engine sponsored ads. Finally, we talk about new work Jim presented at SIGIR studying the search process as learning (not decision making).

Check it out at Free IQ.