MRS Social Media Research conference

The first of its kind happened today. I haven't the time to do a full blow by blow analysis of the conference . It was a privilege to be involved. I was asked to participate in one of the panel discussions.  I have separately blogged about the Andrew Keen keynote.

It felt a bit like the premier league at times – lots of dazzling international players – and wondering about the future of the national game. Clearly London is a social media centre but the brightest minds working on it have travelled to work here! Secondly it was very easy to fall from clarity into detail and noise rather than signal. The best contributions managed to keep goals and methods crystal clear without descending into mind numbing detail which never works well from the platform.

Annelies Verhaeghe of Insites consulting (remember her from Cloud 4 the previous night?)  gave a strong clear presentation about how to ensure that validity is maintained while collecting social media data. The netnography case study she presented about older people and food was a fascinating one which offset sentiment levels from buzz volumes so you could weight the data by how many were participating and posting. She also made an interesting point about the differences you could get from sampling using languages other than English – this gives a much more focussed grouping than the ubiquitous English language which reveals almost nothing about where the speaker might be. Anneliese concluded that representativeness is not possible but sampling is just as vital to balance the data. She also pleaded for a focus on the middle of the long tail where specialised communities very useful for particular projects – would tend to cluster and would be missed by studies that only went for high volume traffice sites or went grubbing in tiny marginal publishers.

Jake Steadman of O2 and Francesco D'Orazio follwed with a casestudy about building a social media toolkit and dashboard for O2. One which could monitor customer concerns in real time. This is no mean feet. It was interesting that the bottleneck now became the client organisation teaching the different stakeholders how to ask the right questions of the system and to correctly interpret the data it was serving up.  Francesco pointed out that the web is variously a mirror, a projection and a construction site – how it functions affects how you interrogate it.  He was reassuringly dismissive of sentiment analysis – they sample and weight it to neutralise its more extreme manifestations – neutral sentiment doesn't help at all!

I can't remember much of the panel discussion I was on – the trouble with being on the spot is that your attention is being fully given to the moment – so note taking the key discusision points proved impossible. I was on with fellow RLFer Doug Dunn and Debi Bester of Proximity, Stephen Roberts of the Guardian and Andrew Reid of Vision critical. All I do remember was my parting shot when I got irritated with the notion that research agencies had to compete using proprietary methods. I suggest that it was absurd to co-create with respondents but not with clients and competitors. We need best practice to create a healthy category and secretive business practices won't increase quality – it will just let black box merchants rip clients off. I was trying to follow Twitter feeds from the front on the ipad so the panel could address comments personally but its more difficult than it looks!

Andrew Keen – see above – took us to the lunch break.  The afternoon commenced with a second panel discussion Ray Pointer having had to cancel presenting his paper. This was on the subject of validity and representativeness and showed the shoals and deeps it was all to easy to fall into.Once quibbling over definitions started it was hard to keep up. 

Doron Meyassed of Promise and Rick Jenner presented a research community for Virgin Media – what was a little bemusing was why this was part of social media – but they claimed they had delivered the openness of a social media environment behind the closed walls of a community. Social media is nothing unless you get it to work in private as well as in public!

This was followed by a Metro newspaper casestudy about young people and the city which strayed into psychogeographic territory – the kind of cities that teenagers imagine for themselves.

Last paper of the day was from Charlie Osmond from Fresh Networks who powered us through to the finish by giving some suggestions about where social media may be going next and the implcations for research (brave man!) For me the most interesting prediction was the backlash against marketing – good point – the pendulum has to begin to swing back at some point towards some notion of private space which brands aren't colonising and insisting on paying for.

And so to close of play. A lot of the pleasure of the day for me was catching up with colleagues Doug, Tom Ewing of Cantar,Paul Hutchings of Kindle and Lisa Ohlin of Insites. Social Media is still so new as a research area that we are having to listen carefully to one another – if we descend into definitions we can soon irritate one another into namecalling.  But it would appear that plenty of companies are making progress in the social media space as a new territory becoming distinct from the rest of online research. Most interestingly client companies aren't about to open up their own borders anytime soon. Social media is something customers do out there. And which employees do in private in the office – when they are allowed. The two won't be meeting in a shared space anytime soon. Cluetrain manifesto got it right over 10 years ago. Marketers still need to be seen to control the conversation.








Twitter Updates
Photo Albums
TV stations JeddahPutna Sihastra Monastery 2011Grindleford Grinch April 2007Planners drinks night No1 BucharestWillie Williams exhibition

Designed by Matthew Pattman