At the risk of repeating myself, consider the following statements:

1. Never before have brands had so much data about their customers or capabilities to analyse this data
2. Never before have we had customers so willing to interact with brands and each other online (social media, user generated content)
3. Yet, we are the least connected that we will ever be...

It is this last statement that really draws pause from me. As marketers we are really at the very early stages of a journey.

Now, whether you buy-in completely to these statements or not,
there are already some trends that are well established - one being the
rise and rise of "text analytics". I can't say that I'm a big fan of the
labels we give these things but I guess we need the shorthand!

From the organisation's point of view, the analysis of structured customer data has become reasonably mature task and whilst many could do it better it is hardly the driving force of differentiation it once was. You only have to attend the latest data analytics conference in your local city and observe the same old case studies to come to this conclusion.

Structured data is the province of traditional CRM and transactions based systems. This data is very good at providing context - WHO, WHAT, HOW, WHEN and WHERE. It tends to address PAST behaviour and interactions with your brand. And it is usually (though not always) good at positively identifying your customers.

Much of the unstructured data that is now being generated through social media has different qualities. By definition it is being VOLUNTEERED directly by your customers or prospects - rather than forced into a contact centre process. It is a very immediate channel - it happening right NOW whether you like it or not. It can be about previous experience with your brand but it far more likely to address brand sentiment and future intent than other data sources. It tends to be more about the WHY rather than the contextual data provided by CRM systems. And when matched with critical structured data such as identity and ratings it can become an invaluable source of insight to the brand. 

You may have also noticed that it is exploding at a great rate of knots - I believe to the point that that it will push the humble customer insight profession into a completely new league.

The down side to this explosion of potential - you need the tools and experience if you are to genuinely ride this wave and not drown in the data.

So what does text analytics cover?

Text analytics is an umbrella term that covers a range of techniques and practices including natural language processing, text mining, relationship extraction, classification and tagging, visualization, modeling and predictive analysis,

In the customer engagement programs that we are conducting with our clients (primarily Net Promoter Programs and branded online communities)  it is this unstructured data in the form of 'verbatims' or open-ended comments in combination with the structured data of identity and ratings that produces the value for the brand.

As the volume of verbatims increases the challenge becomes, for example, how to:

• tag sensibly and consistently through an automated or semi automated process
• identify emerging drivers or trends in the data
• present visually particularly to front line staff or the less analytically inclined in a way conducive to follow-up action

And by the way, the practitioners secret with unstructured data once you have met these challenges... it is a whole lot more fun!