In March U.S.-based text analytics vendor Clarabridge opened an office in the U.K. and recently celebrated it at the British Library in London. Sid Banerjee, the company’s founder and CEO, brought over key members of his team along with representatives of three U.S. clients. He explained that the new office would enable the company to support international clients better and allow expansion into European countries. The three clients then gave detailed presentations on how they use the Clarabridge product to understand their customers better and to use the insights gained to improve business performance.
All had similar messages. They felt forced to support more channels of communication with their customers, especially social media. As a result they were generating much greater volumes of text-based data – email, forms, surveys and social media interactions - but weren’t using it to generate insights apart from through some highly manual processes that provided very limited results. They had automated the analysis process using Clarabridge tools, and all three said the insights they gained far exceeded their expectations, especially as they could tie together insights gained from both structured and unstructured data.
All three, to one degree or another, were advocates of net performer scores (NPS) and said that analyzing free-form text from their surveys enabled them not only to produce the NPS scores but understand better what was driving customers to score them in the ways they were, both positive and negative. The customers said they used these insights to drive improvements across the business. In effect this confirmed my conclusion that the real value of analysis and metrics lies in taking action driven by that analysis. These actions should take the form of improvements in products, processes (especially those related to activities that cross business units), and most importantly the performance of people, either to focus training or to motivate individuals.
Alongside all these positive messages, there is one note of caution or perhaps better described as a learning experience. Each company had bought the product for its own reasons and with its own goals in mind. But again they echoed a message I have heard from other users of text analytics (and speech analytics as well), which is that a company won’t really understand the full benefit of this technology until trying it. This is best illustrated by an example I often use. Many users start by “defining” what they consider to be a complaint, which many do by trying to spot words, phrases and sentiments that they believe customers use when complaining. The first few trials might indeed spot some complaints, but many users I have spoken to discover that the language customers use is not quite what they expected. It is therefore important to have tools that allow users to “look at” the content of their initial complaints and further refine their definitions until they are not only picking out true complaints but can drill down into different classes and types of complaints – for example by product, call center agent, channel of communication or IVR script – and also the severity of complaint – minor, major or threatening to desert to the completion. Down at this level, users I have spoken to say, they see how to get the most benefit from their text analytics product.
With this in mind, Clarabridge announced in its version 4.5 “Tower” release can help apply text analytics to your customer and consumer level analytics. One of the new capabilities it calls “word clouds” which come as part of several new analytics capabilities. Word clouds show users in graphic form the content of text, helping them identify words and phrase they should be zeroing in on. Other new capabilities include easy-to-use dashboards, more advanced visualization options and improved sharing of data and insights between users. The explosion in social media has led Clarabridge to include interfaces to more social media sites, so users can extract data directly from them and one of my recent points of emphasis. The new release comes with support for two more languages, French and Portuguese, with more promised to support the expansion into Europe. Alongside these function enhancements are several new administration features that make the product easier to set up, configure and manage. This release builds on my last analysis of Clarabridge that I see as connecting the dots in finding causal factors contributing to good and bad customer experience.
My recent research into customer analytics shows that adoption of text analytics is slightly lagging behind speech analytics. At the same time, there is no doubt that the explosion of social media is influencing companies’ requirements. Most companies I speak with want at least to monitor what people are saying about them on social media and determine their overall sentiment towards them. Text analytics is the right tool for these tasks, and as the three companies speaking at the Clarabridge launch highlighted, once you get the hang of using it, the benefits to be had are many and varied.
Have you invested in text analytics? I have provided you the focus that it is part of the contact center revolution in 2011, and now it is up to you to act upon. If so please tell us the benefits you have found and we can in return provide you some insightful research on customer analytics and role of text in your processes.
Richard Snow – VP & Research Director