Web Metrics 1.0 (1993-2001): "How many eyeballs you got dude?"
The hype about the internet boils up and then over:
The press is all about:
- Internet penetration ("My nanna just surfed her first porn site!")
- No. of web sites ("Look, I can trade in my Volkswagen for a JCB, all with the click of a mouse!")
Sites are all about:
- Page impressions
- No. of unique visitors
- No. of clickthrus (half of business models being built on ad revenue)
- Consumer $ spent (the other half being built on e-commerce)
N.B. We still use these metrics. They are useful. But they are no longer enough.
Web Metrics 2.0: (2003-2009): "Our users are sooo creative man"
The press is all about:
- No. of blogs (and why they are full of rubbish)
- No. of article on wikipedia (and why they are all full of rubbish)
- No. of members of Friendster/MySpace/Facebook (delete depending on year)
Sites are all about:
- User growth
- % of users who generate their own content (not as many as we thought)
- % of users who consume
- % of users who comment
All this is good stuff and the research by Forrester & Pew is useful. But it seems like it's missing something.
These measures still treat people as individuals interacting with sites. What's missing is the "social" bit of social media. How do these people interact with each other around the media they create? We need to start getting a bit more imaginative about all this.
We may end up drawing more on social network analysis - which has some well-established metrics. But these don't intuitively grab people*.
Bearing in mind that the point of metrics is to allow comparisons (both between groups and across time) that underpin good decisions, what metrics should we be setting our sights on? Are numbers the answer?
*A similar problem afflicts the gini coefficient as a measure of diversity in virtual environments.
Posts
4 comments:
This is such an interesting observation - I sometimes think that the "social" bit of social media is very often the individual's engagement with the toys rather than as you rightly suggest, what happens between people as a result of the toys (apart from finding someone you had a crush on in high school, meeting up, having an affair, leaving your spouse and the family falling apart :)
The absence of a discussion on this stuff is so interesting ... could it be because there isn't as much "socialising" leading to productivity as one would imagine? I'm not sure of the answers but I do know my eyes glaze over when I hear people talking tech specs about their new gadgets or showing off how many "friends" they have on social networking sites...and don't get me started on the erosion of boundaries this 24/7, access all areas of your life on and offline, that this stuff implies...It's Thursday, I needed to get that off my chest...
Very interesting piece - ta - you probably know this is a bug bear of mine and I just came back from a social tools conference in London where I caught myself stopping my presentation and doing my metallica impression "metrics: GOOD, waffle: BAD".. anyway
Agree with Annette, isn't it interesting this isn't being discussed much? So much that I'm convening a meeting in Bristol this pm to look into this and review the possibility of some form of dashboard to track 'community' activity in different networks (which all have different indicators)...
Miguel's stuff rocks - it's based in CoP stuff so is more about the interactions than the bulk metrics of page views or 'growth' (as you highlighted in phase 1 and 2), and as such, with his background calculations, can show some measure of the levels of our sharing, but we're still not getting to 'what' we are sharing - this will always require qualitative reviews/surveys/trad community management stuff I reckon, but I also heard that the Guardian new media team are exploring this area... so I'll hassle that one and keep you informed...
while I'm looking at your reccommended SNA stuff... :)
Since you're asking for comments I'll write here, although this issue's another that's worth picking at length :-). You're pulling a very interesting string of posts.
The SNA metrics have a sourcing problem. Even though they could be very relevant in CoP management, obtaining data from the wild is very difficult (i.e. you can't pull it out of a web board's stats). So you try to get extrapolations that serve similar ends... and use your eyes to see what's going on. Even when you can map interactions automatically, it's kind of hard to separate the social from the practice-derived.
And, as Ed mentions, I still believe relevance metrics are core to understanding what goes on in a community. Relevance changes less than membership (even than core membership: we've had a couple of chances to experiment), which says a lot about the relative value of each analysis for management.
And yes, a Gini coefficient would be a nice synthesis of inequalities. What would you measure? Centrality? Participation?
Best regards,
Miguel
Miguel - I have the Book of Measurements. Will digest and post shortly...
Post a Comment