Is social data the new focus group?
The algorithm takes raw opinions and categorises them, based on their sentiment and topics.
Social media represents the largest dataset of unstructured human expression ever recorded, and it’s growing exponentially by the day. According to Gary King, Harvard Professor and Director of the Institute for Quantitative Social Science, there is now a billion social media posts made every two days.
Although social media listening is a growing interest for product developers and marketers, the industry as a whole has not yet fully embraced using social data for business intelligence, particularly in Australia. This is hardly surprising considering that finding conversations which are relevant to your business efforts can be akin to sifting through individual grains of sand looking for the salt you dropped.
It’s only in recent years that the tools and platforms have become user-friendly and accurate enough to be widely adopted. These platforms not only aggregate the torrent of real-time social data into one place, they also offer analytics capabilities to make it easier for users to gather and interpret valuable insights. This allows us to go beyond social listening and big data volume metrics, to understand target market sentiment, behaviour, and influencers.
Given social data’s capacity to distill consumer insights, the platforms play a role in the space once reserved for primary research methods. In fact, some commentators have even called social media data the “new” focus group. And while I’m not suggesting that social data should replace traditional focus groups, it does offer a valuable point of difference — unsolicited opinions. Without a mediator or questionnaire to prompt, the researcher is able to eliminate respondent bias, which is big news for new product development or product improvements.
There are several platforms currently providing social analytics, but at One Small Step we use Crimson Hexagon. As One Small Step’s social analyst, I use the platform to understand consumer behaviour, brand sentiment and reaction to campaigns.
The algorithm takes raw, unfiltered and unstructured opinions and categorises them, based on their sentiment and topics of that particular conversation. It means that we can understand the topic and our target market better — sentiment, drivers of influence, and how the conversation changes over time. This gives us insights that we then use to make decisions about strategy and drive our creative work.
An example of this in a recent research project we undertook for a major Australian tuna brand. As well as traditional methods of consumer research, we ran a detailed breakdown on the social conversation mentioning “tuna” and found that the major topic drivers had shifted from environmental concerns, to being a part of the new wave of “#fitspo” promoting healthy proteins and wellness.
This insight was a surprise for the client, who had been focusing the lions-share of their communications on environmental messaging. It became clear that their customers’ conversation had changed, and as a result the brand needed to retarget their messaging. They are now in a much stronger place to create successful communications and frame their product R&D.
What the future holds is exciting. Recent studies have shown that computers are now statistically better than humans at photo recognition and analysis. With the rise of Instagram and Snapchat, the potential of deep sentiment analysis on picture and video content presents an entirely new spectrum to social insights and product learnings.