The ability to monitor what people are saying on social media is nothing new. But the ability to monitor what people are “seeing” will completely change this discipline over the next 24 months.
The first time I used a social media listening tool was back in 2008. It was called Alterian SM2 (catchy, right?) and I was actually quite impressed at the time by what it could do.
Over the intervening years I’ve had to opportunity to use a whole range of other tools, developing a good (I hope!) understanding of what social listening can offer brands, and equally what it can’t. If I’ve learnt anything, it’s that social media listening is both a science and an art form.
Of course the tools themselves have also evolved over that time. User interfaces have got sexier (kind of), new platforms like Instagram have been integrated and new data sets such as Facebook Topic Data are, albeit with some teething problems, started to be included.
However, one thing has remaining very much a constant over those 9 years – the tools search on text. Sure, most vendors offer search capability in multiple languages and you (like me) have probably read countless specs discussing advances in NLP and machine learning. But at the end of the day it’s still always text analysis. Or at least it was…
Eyes up sheriff, visual analytics just rode into town. And it means business.
Now I fully appreciate that visual recognition is in no way a new technology in itself. But in terms of a serious, sophisticated offering by major social media listening vendors… it’s pretty nascent.
The more I read and think about this new type of functionality the more ideas and opportunities spring to mind;
- how often does my brand just “appear in the world” compared with a competitor?
- which brands regularly appear alongside mine? (CocaCola and McDonalds?!)
Sure, there are some funky challenges around how data is sourced, especially with Instagram and its pesky #hashtag-based search limitation. But as our understanding grows clients, agencies and vendors will think creatively to get around that.
This opportunity to bring visual content into the scope of what’s analysed feels like a genuine step change. Give it two or three years and I can see a scenario where visual analysis constitutes the lions share of social listening programmes.
Like most social listening tool developments, we’re clearly playing catchup.
How many of you now use Instagram as your primary social network? Posting images and short video clips is just standard user behaviour on social media these days.
For businesses, that visual window into your world gives amazing context of how their brand features. It’s a just a shame that it’s probably going to be a little while until the tools have the ability to accurately add context and meaning to what they find.
Fair enough, even with all this visual capability in place, most people won’t be finding social listening sexy any time soon. But it will deliver a genuine step change into how things are done.
The question is, what will it make of me posting a Snapchat filter like this…
Man? Dog? Dogman? Mandog?
Are you already using visual analysis? What’s working… and what’s not?