Mining social data for real business results

Joanne Carew
By Joanne Carew, ITWeb Cape-based contributor.
Johannesburg, 21 Sept 2018
Mervyn Mooi, Knowledge Integration Dynamics.
Mervyn Mooi, Knowledge Integration Dynamics.

In the past, social analytics was all about sorting social media chatter into 'positive' and 'negative' buckets. This style of sentiment analysis has been applied to everything from advertising and politics to war. But it's no longer enough. Simply measuring the sentiment in a sample of social media posts actually doesn't reveal any truly meaningful information. While this approach could often determine the 'what', 'when', 'where' and 'how', it didn't always explain the 'why'.

But with social business intelligence (BI), all of this changes.

Let's be clear - when we talk about social BI, we're talking about it being used both internally and externally. Internally, social BI helps a business to better understand employees and gain insight into what makes them tick. It also reveals what engages them and what factors contribute towards their performance and how to deliver these incentives, says Chris Ogden, MD at RubiBlue. Similarly, these same tools can be used externally to gain richer insights into your customer. By using data collated from social networks, you can interpret their motivations and needs, using this data as a market intelligence tool.

Brands used to conduct focus groups to get customer insights and understand their perceptions of a product, service or campaign. And they would hold team-building sessions to find out how their employees were feeling. Although these approaches can provide valuable qualitative information, social media is more relevant because the sentiment is shared subconsciously, making it a much more honest way to make sense of consumer and employee behaviour in today's data-driven world, says Wendy Tayler, social and innovation lead and a director at FleishmanHillard. Social BI allows businesses to make better business decisions because they have a solid grasp on the attitudes and behaviours of their employees and target audience.

Social BI is something businesses of all sizes should consider, notes Accenture's MD for Technology Strategy, Hans Zachar. For the small business, it provides an easy opportunity to learn more about your team. This information also makes it possible to leapfrog the big guys and provides a level of exposure, and a means of consumer engagement, that traditional marketing could only hope for. Today, larger brand are simply expected to have a social leaning, dovetailed seamlessly with all customer-related interactions, communication channels and internal business processes.

The right tools

For Taylor, getting this right means removing emotions from your analysis. "Sometimes data doesn't display the information that you expected or wanted to see, but the point of BI is that it is based on objective evidence to help you make business choices, not emotional ones." She also believes it is essential to invest in the right tools. With so many analytical and social listening tools to choose from - many of which are extremely expensive - she suggests that businesses do their research and pick a solution that suits their needs best. Or better yet, consult with an external party to understand what options are scalable, she adds.

Sometimes data doesn't display the information that you expected or wanted to see, but the point of BI is that it is based on objective evidence to help you make business choices, not emotional ones.

Wendy Tayler, FleishmanHillard

By starting with simple social listening tools, you can quickly gauge the sentiment in your market and then respond to that sentiment, advises Marc Gower, Dynamics lead at Microsoft SA. This approach offers the quickest wins and returns on your initial investment. There is a perception in the market that this process is complex, but there are a variety of tools that one can implement and can deliver actionable social BI insights within days of being up and running. In line with this, companies must remember that the data is already in the network, notes Gower. Many organisations hold back on their social BI implementation because they feel they need to start off by looking at the data on their own social networks first so that they can understand the process before they make an investment. He believes that businesses should make it a parallel exercise - analysing their own corporate social data, as well as data that has been retrieved from external social BI mining.

But businesses shouldn't be complacent. This is a challenging discipline, and certainly won't be a walk in the park to implement, advises Julian Thomas, a principle consultant at PBT Group. The key is to find a clear problem or use case in the business and then identify a good person to partner with, someone who will walk the journey with you. Everything you do must be tied back to the key strategic objectives of the business, which is why you must understand the business value chain and the main business drivers. Like Gower, Thomas recommends that businesses start small. Once you have a use case, you'll get business support and then it's a lot easier to build momentum and start ramping up on that capability.


Unpacking the trend

So, is social BI the missing link needed to properly meet customer demands?

The experts are divided on this one.

According to Gower, while social BI does provide business with market insight, the organisation still needs to rethink its go-to-market efforts and then redefine and redesign this strategy. Sure, having these social BI insights highlights what changes are necessary in order for you to grow your market share, but to really change customer experiences, you need to use the insights to transform your entire organisation.

On the one hand, a business can reveal customer preferences and sentiments using social media and then adjust their products/services to meet these requirements, notes Mervyn Mooi, a director at Knowledge Integration Dynamics. But businesses can still gauge customer requirements using more traditional methods and CRM systems.

Does making the most of social insights require a mix of tech and human touch?

Because most social insights are subjective, we need to get people involved to make manual adjustments and final decisions, says Mooi. We know that technology can automate derivations and generate results to a point. And it can also do decision-making to a certain extent. But it is humans who control the end result. This is likely to change once artificial intelligence is used more widely, but for now, people are still a must.

The technology allows us to analyse and aggregate the social data, but deriving insights from this data requires human interpretation, with context and nuances overlaid, adds Taylor.

With close to half of the world's population plugged into social media, are companies making meaningful use of all this data?

Companies have become very good, or at least not terrible, at managing unstructured data, says Zachar. Social media data is unstructured by nature. This means that if companies want to make meaningful use of this data, they must transform data sets into something where trends can be analysed, unlikely linkages identified and plans can be rolled out to address causal relationships measured. A lot of thinking is going into how to marry the world of the data scientist with that of the business owner and truly transform the marketing function into an analytics-led organisation, he continues.

One of the greatest challenges, says Zachar, is the very practical one of timing. Social media data is instant and real-time, while operational and traditional marketing data typically comes with a lag. Customer expectations are that their social media conversations are instantaneous. There is little tolerance for organisations that respond several hours or even days, later when the operational data stores eventually catch up.

Today, plugging into social to better understand your market and your employees is a non-negotiable for any company. But there isn't - at least not yet - a one-size-fits-all model available. Different clients and different industries will have different requirements. Finding the right fit is all about asking what levels of scope and sophistication you need from these platforms in order to successfully manage all of this information.

The local customer conundrum

While developed markets are moving swiftly into fully digital economies, South African consumers pose a challenge in that they still largely transact physically, but tend to complain online, says Accenture's Hans Zachar.

The issue with living in this blurred journey from physical to digital is that companies tend to leave the traditional business as-is, while they develop capabilities to service customers digitally in a largely separate way. In this scenario, too little focus is given to integration between the two. As a result, it becomes very difficult to track customer interactions, leaving the digitally savvy customer, with much higher expectations and much less patience, unable to understand why the company can't resolve issues based on his or her emotional Facebook post. "Far too often, social media is used as a name and shame mechanism when, in fact, it can become a legitimate interaction and engagement channel."

Obstacles ahead

For Julian Thomas, of PBT Group, it all comes down to discovering what is going on in the minds of your employees and customers. Some of the main challenges that need to be addressed include:

  • Being able to gain access to sufficient social data - freely available data vs private or protected data. And once you have this information, being sure to comply with legislation like the Protection of Personal Information Act (PoPI).
  • Making sense of the data - this is where understanding the `why' of customer and employee behaviour becomes critical.
  • The cost of accessing data - such as paid-for media sources, for example.
  • The human challenge - having people in an organisation, in both business and IT, who are able to use this data to identify opportunities and convince business to act on these insights.

This article was first published in the September 2018 edition of ITWeb Brainstorm magazine. To read more, go to the Brainstorm website.