Riding the wave of social conversations

Brands are just coming of age in the new collaborative economy.

Like any teenager though, a lot of their actions at the moment are a bit awkward. Through those uncomfortable times, most of us learnt at some point that joining in on an existing conversation is much easier than starting your own by shouting until you get someone to pay attention. But to join in and be heard means being relevant and culturally on code, and of course the ‘on code’ bit is what brands need to work out because it is constantly changing.

We know from years of analysing anonymous and aggregated social media conversations through Zavy that people mostly talk about micro-trends on their social feeds. But most micro-trends can last anywhere from a few days to a few months, so the problem for brands is largely one of timing. With trends that become old news this quickly, the model of insights generation through research and marketing communications development is far too slow. Having a survey in field one month to getting results the next, and then spending several weeks developing a marketing response will not deliver cultural relevance – a new approach is needed.

These micro-trends are emerging and fading all the time, so brands need to be able to identify and act on them at the right point in their cycle. Too early risks looking like sponsored activity by the brand, too late and the brand seems like that friend who gets the joke just a little late. Monitoring the collective conversation in digital and social channels makes real-time insights possible, which in turn means that it's realistic for brands to hop on the wave of a micro-trend at the right time.

Analysing digital data to find cultural trends is not a new idea. Google has had a service called Google Trends available for nearly a decade and has produced an annual zeitgeist report for longer than that. Ideas like tracking the spread of the flu using Twitter data or Google searches is also a much-discussed topic. And we too have been tracking macro and micro trends through Zavy for the last 3 years. Two things have changed recently though that make this an increasingly interesting space.

Firstly, the type of data that can inform this cultural analysis has exploded with the growth in social media channels and the increase in availability of that data from the channels themselves.

Secondly, and more importantly, our analytics techniques are finally catching up to the massive amounts of text, image and video data now available. Machine learning as a field has undergone a renaissance in the last few years and 2016 was the peak of this growth. Billions of dollars of investment were poured into new startups in this space and the big industry players like Facebook and Google open sourced some of their core algorithms. This makes extracting information from the stream of images, text and video data flowing through the social space feasible. While aggregating this data and identifying trends is not easy, it is at least possible.

Perhaps the most exciting development is cracking image recognition and coding. Just a year ago this was difficult and low quality, but now it is easy and very sensitive. We know that the visual sense for many people is their dominant sense, as evidenced in the recent rise of video material. So it is that in many markets micro-trends emerge through images in the first instance, followed only later by a written articulation of the ideas through posts and blogs. A blend of image capture analysis and qualitative coding can capture a movement in a market that is still nascent but nevertheless widespread and emerging.

These developments have been inspired by Paul Lester’s work on the syntactic theory of visual communication: “We are becoming a visually mediated society. For many, understanding the world is being accomplished not through words, but by reading images”. Seizing the opportunity offered by new technology, a European-based food company recently completed an international segmentation which successfully directed product development based entirely on an analysis of food imagery on social media.

Understanding culture at a point in history is not enough, we must have an ear to the ground and know what conversations are happening right now if we want our brand to be relevant. Grown up brands are increasingly getting this and now we have the tools to make it happen.