What interests me is that there is a qualitative aspect to the coverage. Take for instance this entry on a recent paper:
If you have the bookmarklet you can track down who is saying what. That’s much more value than a number because it shows where the debate around a paper is going. I think this is why directly gaming altmetrics will be difficult. Easy solutions will add little value. For example, I could go to fiverr.com and pay $5 for someone to retweet links to our papers. However I don’t see what bland links from accounts with no followers will add qualitatively to the conversation around a paper, and I think it would show up pretty easily in the altmetrics.
I’m looking improving how we link to papers, our own and others, for a few months. Today, I’ve released a prototype plugin today for WordPress blogs (not wordpress.com blogs, sorry) that handles linking to DOIs, figshare, arXiv and ISBNs. It produces a citation that I hope ScienceSeeker.org will be able to read and a META tag to help sites like altmetric.com. If you want to try the beta, you can pop across to my personal site and download the plugin to test yourself. Once I’m happy it’s not going to accidentally destroy anyone’s site it’ll be released with a GPL, so anyone can use or re-write it for free.
The target is to produce something that adds value by making connections that are already there easier to spot, and not creating spurious mentions. Of course, I might have a lower standard for value than you. Altmetric.com’s post is a helpful starting point in working out what we should be tracking.