Klout and influence

There was an interesting discussion on Twitter yesterday between Charlie Stross and Pat Cadigan about Klout.com, which I vaguely knew of as one of those sites that attempts to measure how wide and effective your use of social media is, attributing to me expertise and authority on subjects about which I know little and tweet less. Klout had further annoyed me by unilaterally changing their algorithm to decrease my Klout score and then demanding that I change my behaviour to increase it back again, which didn't interest me at all.

@cstross made the excellent point that Klout's data-mining of our Twitter and Facebook accounts is ethically dubious and its opt-out rather than opt-in practice is actually against EU law (ah, how I remember the days of lobbying the European Parliament on that one), and that was sufficient to move me to opt out.

Later in the day, Northern Ireland's Social Democratic and Labour Party elected its new leader. Delegates chose the Member of Parliament (and of the Northern Ireland Assembly) for South Belfast, Alasdair McDonnell, whose candidacy I had endorsed in a blog post on Wednesday (reprinted on the popular site Slugger O'Toole on Thursday). I doubt that more than a couple of dozen SDLP delegates read my piece in either place, and I would be very surprised if I changed the minds of more than a very few (and McDonnell's winning margin was a healthy 36 votes out of almost 350). I can tell you that Livejournal thinks that 75 people looked at it as their point of entry, compared with 53 for the preceding entry about the beards of Democratic Party candidates for the US Presidency and 149 for the following entry about Tintin. It was tweeted by two people who both have fewer followers on Twitter than I do, but probably are more connected to the Northern Irish social media world than I am – one is the communications officer of a prominent Belfast NGO, and the other is, er, Alasdair McDonnell MP. None of the other half-dozen pieces on Slugger about the SDLP leadership attracted fewer comments than mine.

I think Klout – or any naive data-mining algorithm – would have real difficulty in assessing whether or not that piece was influential, let alone whether I am influential, just from the numbers. To give another (and rather appropriate) example, my re-posting of the graphic of Where Should You Post Your Status?, which had been doing the rounds on Facebook, got 100 distinct views on Livejournal, but was also tweeted by five other people (and retweeted by a sixth), only one of whom I actually know, so the others put it into completely new networks, thus increasing my "deep reach" for that particular post (which however is probably of less historical import than the selection of the new leader of the SDLP). For another example of "deep reach", my reaction to the Stross/Cadigan discussion caught the attention of Rosi Sexton, whose twitter following probably doesn't have a lot of overlap with mine, let alone the two sf authors', and so the Klout discussion is brought to the attention of a whole new audience. But I suspect that the "deep reach" of a single post is normally pretty ephemeral; even if it can be measured, it is only a very small part of a bigger picture.

I like looking at attempts to tease structure out of the sea of information that we are all providing online. But I think any attempts to reduce one person’s impact to a single number should be treated with suspicion, and further efforts to make money out of such a dubious process should be treated with disdain and shunning.