Understanding the Value in Natural Language Data

From speech to text, to websites, wearables and even social media, the number of ways customers can interact with an organization is growing rapidly each year. Natural language is a powerful thing, but it’s easy to miss the details that can make a difference to a business and understanding its true value takes more than simple analytics.

Some analytic systems just look for a key word or phrase such as ‘failed delivery’ and check the tick box for that category. There’s no recognition that the customer was actually thanking a company after experiencing delivery problems with a competitor. Others perhaps understand the sentiment, making the analysis more correct, but can still only deliver reports that contain predetermined information.

The problem is that for most organizations the information required for this week’s set of business challenges, won’t be the same as next week and predicting the following month involves some serious guesswork that is often dependent on which way the board decides to focus. Businesses no longer operate on a static level, but the data they are using to drive it is.

Even when more sophisticated analytic systems are used, they take too long to produce useful reports. That means that this week’s data analysis was actually taken maybe as long as eight weeks away – frustrating for most businesses and a nightmare for anyone influenced by seasonal trends.

Even more frustrating when one realizes that just a small increase in sales can transform the bottom line for many organizations. But it’s not possible to make a serious difference on old data, and waiting weeks or even days for actionable data is not good for business. In order to see why it’s probably best to look at some of the more interesting ways some companies are using their unstructured customer data.

Recognizing that its customers felt compelled to contact them, even after they had placed an item in the online shopping basket and continued to purchase it after the contact indicating that there was nothing wrong with their choice, one business analyzed its live chat and contact form files to find out why. After identifying that the customer required more detailed information to be certain they were purchasing the right item, the organizations then used the analyses to discover exactly which products required additional online descriptions and what information needed to include, ultimately reducing the number of inbound enquiries to its contact centre.

Another organization is more proactive in obtaining its data, demonstrating how speed in getting the results is crucial. It instructs its online virtual assistant to ask certain questions of customers when the opportunity arises naturally during a conversation. It then uses this information to consider future subscription services and other product development.

But getting value out of your unstructured data resources is not just about finding out the right information at the right time, it’s about being able to slice and dice the data in different ways that suits an organization to extend its reach even further. Customer service may be interested in product issues, sales might be concerned about purchasing trends and marketing interested in short-term influences. All the answers are there they just need to be discovered.

For instance Google analytics might show what links a potential customer clicked on or pages they visited, but it doesn’t show why they didn’t purchase. This type of information such as wrong size or lack of a particular feature is hidden in search queries or online chat logs. The information is there and it may well eventually be discovered, but perhaps not until two months later when a competitor is already offering something similar in just the color hundreds of customers wanted.

Whilst conventional analytics may still have a part to play within an organization, to maximize the benefit of the data that customers give freely everyday requires a near real-time intelligent understanding of natural language conversations, and all the intricacies they entail. Without it, you’ve missed the entrance to entrance your customers.

Andy, who lives with his family in the UK, is Chief Marketing & Strategy Officer at Artificial Solutions. A regular speaker at industry conferences and events, Andy delivers insight on the rise of AI, the challenges businesses face and the future of intelligent conversational applications.

Leave a Reply

Your email address will not be published. Required fields are marked *