Getting actionable insight from social data

Whatever you call it – social media listening, online analytics, social media intelligence – knowing what is being said about your brand on social channels is a great start. However, brands that simply capture the data will soon find themselves exposed: the real skill lays in the identification of the right data sources – not all data sources are created equal – and the interpretation of the data to identify actionable insights.
Few brands and organisations exploit social media’s derived insights to their full potential. To give a recent example, in the Scottish referendum, the ‘Yes’ camp clearly dominated social conversations. Yet, this insight alone was a poor indicator of the final result.
In the run up to election day, we ran a big data experiment to see if by using a variety of social and other big data signals we could offer our own prediction for the outcome. Arguably, this was a unique event but it provided the perfect setting to demonstrate the potential wealth of data and insight available to brands to help bring together a big picture outlook.
So how can brands, regardless of size, tap into something that has the potential to offer significant competitive advantage? There are a few points to consider.

Audience balance
The more channels monitored, the more rounded the overall view of conversation. Having said that, simply monitoring all channels without considering their value to the brand is ill advised. Brands must choose each channel carefully based on relevance, but be aware that each will offer its own unique imbalance.
Tone and content between social networks and channels can vary significantly even when identical topics are being discussed. For example, when a travel brand gets things wrong its disgruntled customers won’t hesitate to air their views on Twitter. When they get it right on the other hand, satisfied customers might eschew Twitter in favour of leaving a positive review on TripAdvisor, for example, where the brand’s target audience might be more relevant.
This must be reflected in the campaign analysis by applying a greater weighting to the most significant and relevant channels in order to recognise post value.
As a result, simple numbers, such as the number of likes or retweets, should not drive social KPIs. It is highly satisfying to see a post shared thousands of times but if this reaches the wrong audience, it fails to deliver on its objective. Instead, postings within significant and relevant audience channels will return far greater benefits than multiple posts across a small and uninterested user base. The old adage, quality is better than quantity, rings true.

360° approach
Social media is only one small part of the customer picture and it does not represent the full reality. It delivers a skewed view of the brand and organisation. The challenge is to create the best mix of data sources to get the right results. By combining social data with other big data sources, such as customer renewal figures, customer service communications (calls / email), lapsed customers etc. brands can put social media activity into perspective.
In other words, if customers vent their frustrations online but the number of renewals or repeat purchases remains steady then a brand might simply need to improve its social media customer service to neutralise any negative voices. If, on the other hand, there is a correlation between social media complaints and lost customers, more rigorous changes at business level might be needed.
Combining social media with other internal and external data sources allows brands to identify the underlying challenges and opportunities and create business advantage in the process.
When volume is a campaign target, underlying trend changes are much more insightful and valuable to feed into any big picture analysis. Spotting ‘hot topics’ or shifting trends can often represent an exciting opportunity for clients to better understand their ever-changing consumer landscape. When this insight is fed back into the organisation it can help inform product development, marketing campaigns or customer service programmes.

Unique challenges
There is no one size fits all methodology and every client, market and industry presents individual and unique challenges. A multi-channel fashion retailer for example will capture customer data from a number of customer touch points – in-store, online, mail order, call centre to name just a few – and these will all need to be reflected and taken into account when analysing social data.
Every brand and every audience is unique and to get the most out of social media, bespoke audience analysis will provide the best possible insights.

Practice makes perfect
As with all models, continuous learning helps shape better interpretation. When brands are not receiving the desired results, rather than being put off or completely change tack, they must make small changes at every step of the way. Short-term results are great, but sustained analysis will offer additional value in the mid to long term.
Social analysis, especially when combined with big data, should be seen as part of an extended customer strategy and not a quick win scenario for instant returns. By focusing on social data alone, brands are more likely to end up with a limited or skewed interpretation of actual events, issues and opportunities. Instead, brands should take the opportunity to combine social data with other data sources to better interpret audience intentions and drive more informed decisions.
Oh, and in case you wondered, our big data experiment yielded the right result: proof that you can combine big data and social media data to accurately predict certain outcomes.

Claire Snook is content strategist at sixth sense