What I said at Hello Culture 2014

I went to the morning session of Hello Culture, a one-day conference discussing ‘big data’ in the context of arts and culture. I was on a panel called ‘Data – Is the Tail Wagging the Dog?’

I was given a few minutes to talk to the theme and so I put some slides together and wrote a few notes to accompany. This is almost as I delivered it but this version has an extra slide:




I thought I would say something about making use of big data and social media.

Obviously ‘listening’, to big data implies collecting, analysing and creating new knowledge from that analysis. What you do with that knowledge is up to you but I used this picture, from the german film The Lives of Others, to just put a note of caution in. A report from June 2013 by ComRes found that 68% of UK respondents were concerned about their personal privacy online. And here I am about to go through how to scrape personal data and create value from it. You should remember that people aren’t on social media so that you can harvest information about them. Just like in the film they might have an awareness that they could be listened to, but they don’t actually know you’re doing it.

But anyway, onward. There are the questions that I thought relevant to how we might think about the data that comes about as a result of people using social media.

  • Where’s the big data that tells me what people say about their everyday cultural engagement?
  • How do I capture it and make sense of it?
  • How do I create value from it?

You have to realise that the way in which we use social media creates an immense amount of data. The internet is full of scary infographics telling you that there’s a million tweets a second or some such nonsense. None of this is helpful to you. Please look away from the infographic. Your audience is not the entire world. Better instead to think about big data in small places. Maybe the small places in which your cultural organisation operates.

By way of example here’s some recent research I did amount the level of social media activity use, in one month, on the B31 Voices community news service. There’s still a lot of data here but it’s on a level that’s a bit more manageable. So really I want to make the point that ‘Big’ social data is overwhelming but ‘biggish’ social data, relevant to your interests, is manageable.

I’ll come back to this B31 data in a moment but first wanted to say that I think there’s two ways to deal with social media data. The first is about listening for what people say about you because you want to immediately react to it. So a custom search, as in this example, in tweetdeck, let’s you create a narrow search that you can monitor on an ongoing basis. You do this because you want to react to the data as it happens.

Another example, like this is using an app like Loci which finds geo-located tweets, facebook or instgram updates. I like this one as it includes pictures. Big data isn’t just words, it’s pictures too. In fact making sense of image-based big data is something that researchers are only just focusing on.

The second way is perhaps a more reflective, analytic way. That is, to scrape data, using the Google or Twitter API. Having a sense of what APIs let you do is really important. You should have a play with the facebook graph api. It really is not rocket science although this screen of JSON code may suggest otherwise.

What you really want is data in a spreadsheet. Then you can do some analysis and start to reflect on what it tells you. For this, pivot tables are your friend. So for the B31 Voices website everyone wants to talk about Pets. Missing pets, found pets, cute pets, dead pets. Obviously I’m not specifically researching the arts conversation but it’s there.

And what are people saying about the arts in B31? Well it’s not always what you want to hear but it may be that your organisation needs to be aware of it and hear it.

Finally I just want to highlight one of the ways in which to create value from this data. No doubt we’ll hear today about profiling and targeting audiences through data but don’t forget that the data is qualitative and has value in and of itself. B31 Voices turn their data into an attempt to tell positive stories about the people and places of B31. Sometimes it’s worth taking big data off the spreadsheet and making it work for you. ‘Big Data’ is really small qualitative data in disguise.