Joe Rice, Head of Growth & Strategic Partnerships, Twitter
What’s happening? These two simple words neatly articulate Twitter’s purpose and in 2016 became the Twitter tagline. And while Twitter has always been known as a place for breaking news, perhaps its greatest strength is the insight it provides into the mundane. It’s in the millions of opinions shared on anything and everything, from feedback on products and services to people’s opinions, moods, and mindsets. To market researchers, these are insight gold.
Like most professions, market research and consumer insights gathering have undergone tremendous, technology-enabled change over the past few years, however, most of that innovation has focused on improving traditional primary research methodologies, namely the survey. While traditional surveys undoubtedly provide great value, adding the powerful information that comes from the unprompted and from observations of what’s being said in the wild, ‘organically’- has been a major opportunity Ipsos and Synthesio have seized (more on this in our upcoming webinar). Unsolicited consumer commentary shared on social media remains vastly underutilized by the industry and yet represents a quick, cost-effective way of understanding consumer trends and preferences in real-time. And, by amplifying the literal voice of the customer, social data does so in an authentic way, with a richness and emotion that solicited methods like surveys often lack.
So what is holding the industry back from fully adopting social data and the valuable unsolicited insights it provides? Common criticisms have often included issues with data quality, representativeness, and sample size. These have some validity, however, we shouldn’t forget that these are challenges common across many data collection techniques – and social data is often best used to spot early trends or take the general pulse of a population. All approaches have their strengths and weaknesses, none are foolproof.
To be clear, social intelligence shouldn’t be thought of as a replacement, but rather as a complement to other methods, and in most cases a great starting point to the research process. In fact, I’d argue that social data’s killer app is in the exploratory research stage where the focus is on obtaining a high-level overview of a topic and the general themes associated with it. This allows initial hypotheses to be formed and can even aid research design. At this stage, you are not concerned with conclusive results and therefore some of the aforementioned challenges of social data aren’t as relevant.
Another barrier to utilizing social data has been the challenge of analyzing large volumes of unstructured text. This isn’t a dissimilar challenge to that of coding open-ended survey questions, which has historically been labor-intensive for research professionals. The emergence of artificial intelligence (AI) and, most recently, the incredible improvements witnessed in Natural Language Processing (NLP), have transformed our ability to categorize conversations and identify consumer-defined themes on a massive scale. Synthesio’s new Topic Modeling solution is a great example of this innovation.
The final challenge holding social intelligence back has been its accessibility. For most of social data’s history, the only way to find relevant content was via the creation of Boolean queries allowing you to filter through millions of social conversations. And because the user didn’t typically know what exactly they were looking for (because they either didn’t know what they didn’t know or didn’t know the words and phrases used by social media users), the process of creating these queries was a mix of trial and error. Here again, AI can help by automatically annotating social posts with multiple classifiers, thus expanding the quantity and quality of social conversations connected to a given topic. These are conversations that would have otherwise been missed as they didn’t include the exact keywords. This move beyond Boolean should expand social intelligence to a whole new group of users who can leverage it in innovative ways unlocking new insights and use cases.
It’s now been 16 years since Jack Dorsey sent the first Tweet signaling the beginning of a public conversation that has grown to encompass hundreds of millions of people around the world who share their thoughts, experiences, and opinions as they happen. What was once viewed primarily as a quantitative data source focused on engagement metrics and top-line volumetric summaries, is now becoming valued for its qualitative side and the richness of consumer insight it provides. Not every research question or business problem can be answered using social data, yet its ability to provide a unique window into people’s lives is incredibly powerful and deserves more attention. If you truly want to understand What’s happening, Twitter is a great place to start.
To learn more about the role social platforms like Twitter play in market research and insights – and how blending data types like social and survey can help brands understand consumer needs and behaviors – register for our upcoming webinar on September 15th.
