The rate of social media usage and adoption continues to expand but it still may only represent a small percentage of your total consumer audience. The bulk of your business intelligence may actually be wedged and embedded in other sources. But it isn’t simply an either or situation, where your business has to choose which data to analyze. In many ways, social media and other private data are two variants of unstructured data.
The rate of social media usage and adoption continues to expand but it still may only represent a small percentage of your total consumer audience. The bulk of your business intelligence may actually be wedged and embedded in other sources. But it isn’t simply an either or situation, where your business has to choose which data to analyze. In many ways, social media and other private data are two variants of unstructured data.
Social media conversations are unsolicited, real-time expressions of consumer preference, intentions and opinions. Private data, depending on its source, can represent organized text exchanges between team members, solicited responses to survey or issue-resolving customer service chat transcripts. In both cases, the text is free-form, inchoate and maybe even a bit disorganized. The point is insights into consumer preferences, issues, ideas for product development are expressed in a format that does not lend itself to easily analyzed.
Ideally, the best approach would be one that can treat and analyze unstructured text in the same manner regardless of its source. Consistency in how data is managed and organized is critical if the results are to be correlated or integrated with more traditional data sources. An example I like is one that describes a customer using traditional email or phone to lodge a complaint but failing to achieve any success turns to social media to vent their frustration. There’s a relationship between the initial request for help via a traditional channel and the subsequent use of social media; the connections are the:
- issue the customer was calling about
- the product or service
- the customer
- the sentiments the customer felt navigating from one platform to another
The initial trigger for the issue may be:
- an upgrade
- problems with an existing system
- an outage
The big question then is how to begin to isolate the attributes of these conversations that may be taking place within different platforms and settings and make the connection. What if you were able to analyze customer service chats and then correlate those results to activity levels across the social media platform? But the goal wouldn’t necessarily be simply to track volume levels but rather the context and content of those fluctuations. In other words, let’s say you could monitor data exchanged between consumers and customer service reps during an unusually high call volume time-frame and then track those results to consumer comments expressed on social media. Would there be a relationship? What sort of information would be surfaced? Are the same topics or issues expressed in both settings? Sometimes it’s less about the volume of activity or mentions on one platform and more about the relationship between events both social and other private data, the context, and the timing that may be driving discussion.
