Steve Rubel over at Micropersuasion says measuring influence via online links is dead and new factors should be considered. While some of his argument makes sense, it's not exactly new thinking--check out the group think led by Mary Hodder back in 2005.

I agree influence can't be calculated purely through link counts. After all, influence is about the context of what's being discussed and the measurable behavior of audiences that pay attention to those doing the influencing on that topic. For a huge amount of content on the Web, linking behavior is a very important metric, because it implies that someone else (besides the content creator) finds value in the post. The originating content, site popularity, and timing of links are all telling metrics. By linking to your content Steve, I am saying it has some value to me, and pass that notion to others whose readership and trust I enjoy. If others choose to link to your content, that grows its value even more. So how wise is it to be dismissive of this behavior?

I'm also intrigued by Steve's assertions that the number of friends I have on Facebook (10) or the LinkedIn connections I have (145) factor into my influence about a given topic. It's hard to see past those numbers being a proxy of my general popularity. If Steve's right, should I take time to be solicitous of friends in those environments? And since I'm not Twittering just yet, does this mean I'm somehow less influential?

Bottom line: Think what you may about links, but it takes a number of implicit and explicit factors to establish influence. And challenges of using some or all of them aside, if we can't attach measures of audience behavior to the metrics we produce, advertisers PR practitioners and publishers themselves will never be at ease. And if the ultimate goal is to measure influence as a means to monetize, we need mechanisms that reflect crystal clear ROI.