In Search of Actionable Insights from Social Media Data

Most people recognize the incredible potential of the internet and social media data to collect and elicit opinions about millions of topics from millions of people. Instead of using traditional market research methods to seek out target audiences and ask them specific questions, one need only search for opinions in the social media space.

Most people recognize the incredible potential of the internet and social media data to collect and elicit opinions about millions of topics from millions of people. Instead of using traditional market research methods to seek out target audiences and ask them specific questions, one need only search for opinions in the social media space. Even better, there are myriad tools available to aggregate and summarize those opinions for you. Unfortunately, I consistently hear one common complaint – the data and the results are not actionable.

In many cases, these complaints are unwarranted and avoidable. Let me take you through three scenarios.

1) The data truly are not actionable.  The social media aggregation tool has collected every piece of information into one single space but failed to bucket, or worse properly bucket, the data into usable sections. Without additional manual intervention, the data cannot be easily separated into meaningful units, such as product colours, product features, or product flaws. Social media data products that require hours and days of manual work are about as good as manually searching the internet. Not actionable? Agreed. Drop the tool immediately and move on.

2) The data are actionable but you aren’t an experienced data user. This second scenario is far more common as social media data often lands in the lap of people who have little or no experience with market research data such as surveys or focus groups. Even in cases where the data has been bucketed into product features such as wifi, coupons, colours, shapes, and flavours, it can be very difficult to see how “positive sentiment towards red” and “negative sentiment towards cinnamon” is an actionable result. Not actionable? Disagree. You have two options here. First, THINK HARD, don’t be lazy, don’t think it’s too difficult, don’t think it’s impossible, dig into the data and stir up those brain waves. Second, find a trusted and experienced market research professional to help you with the data. This is what they do all day, every day, and they know how to do it.

3) The data are actionable and you are an experienced data user. Here we have the intersection of two wonderful worlds. First, we have a social media data aggregation tool that validly buckets comments, status updates, and opinions into clear groups. Second, we have data users who have years of experience interpreting “I prefer this product in red” and “I disliked the cinnamon flavour” for clients.  When these two attributes join together, there is no such thing as unactionable data. Every piece of data is in line to become an actionable insight.