Tuning into the Collective Voice of Our Customers
Putting the customer first is one of TIAA-CREF’s core values. However, without the proper tools and expertise, the cacophony of customer feedback can sound like static coming from every direction. Customer’s communication preferences span in-person, phone, mail and digital channels, and finding the real value in the feedback is a daunting task. Companies must have the tools to aggregate and analyze customer feedback across all channels to truly gain value from this rich data source. Further, they must field a competent data science team that can analyze the complex and volatile signals from both within and outside the company, and distill them into clear, actionable insights. TIAA-CREF, a financial services company dedicated to serving people who serve others, has just such technology and just such a team.
New technologies for handling and analyzing high-volume, fast moving datasets are now making it possible to take this comprehensive approach to customer interaction analysis. These platforms are able to handle complex unstructured data as well as traditional structured data, making them perfectly suited for the analysis of customer communications. At TIAA-CREF, we have taken full advantage of these capabilities to devise tools that will allow us to tune in to the collective voice of our customers. TIAA-CREF has taken the additional step of standing-up an Analytics Center of Excellence to leverage these data science tools to develop solutions to business problems.
The data science based solutions being developed combine feedback across all channels, and sources, in near real-time. Customer-facing staff can now be equipped with the most relevant topics and client comments on-demand. The new data products in the arsenal empower advisors, managers and analysts to proactively address emerging customer issues, while strategically responding to customer concerns.
TIAA-CREF’s Analytics Center of Excellence utilizes fast-fail discovery techniques to dive into customer feedback data and provide front-line staff with more actionable insights. Data scientists are embedded in the lines of business to tackle a pressing business concern with analytical and machine learning techniques. The insights discovered by the cross-functional team have helped create a repository of reusable predictive data which has been leveraged in subsequent engagements across the business. The team uses their enterprise analytics toolbox and expansive capabilities to convert unstructured data into insights over several weeks, outpacing traditional analytics initiatives.
One application created from this effort includes the ability to display near real-time customer comments and feedback. It identifies emerging trends and provides search capabilities for technical and non-technical users to quickly find relevant customer interactions and context. Additionally, the application answers the question “What is the expected impact to my business, based on current trends in our customer voice?” The team focuses on visualizing the insights, and designing an experience that functions across all device mediums, both mobile and plugged-in.
The insights delivered so far have centered on trend identification instead of event frequency. The tool distributes insights that highlight which topics are currently trending, as compared to the normal flow and tenor of feedback. This new approach reduces the noise across millions of events being monitored to a manageable, actionable list that the business can react to.
Using sophisticated text analytics and trending algorithms, the team ingested multiple unstructured text data sources and executed the analytics at near real-time. The text analytics involved include terminology standardization, hashtag implementation, natural language processing, and pattern matching. With the data structured, the application calculates trends using an in-house developed trending algorithm created by the data scientists. The algorithm identifies words and phrases that are being used more now than they have in the past, similar to the way trending topics are reported by Twitter. The solution can rank trends based on their importance, and provides insight by near real-time trends, emerging trends and persisting trends.
With many of our front-line customer interactions taking place from our national contact center, this service area is in many ways the face of the company. The ongoing nature of this customer centric work is one of the key drivers for this timely, automated and proactive method to better understand the “voice of the customer” and to ultimately understand why customers are calling. This new application is a game-changer in that regard. Before its development, call center directors, team managers and advisors were limited when trying to understand why clients were calling and what their top concerns were. “This empowers our customer-facing employees to see what’s trending across each channel, gain insight into what’s important, identify where risks exist, and understand how to better serve customers,” adds Joseph Sieczkowski, Chief Architect. TIAA-CREF is fundamentally changing how it listens to and proactively addresses client needs in near real-time.
By deploying advanced data analysis technologies and by fielding a strong data science team TIAA-CREF keeps pace with the rapidly changing stream of customer communication. Technology executives need to recognize the value that advanced analytic capabilities can drive and ensure ongoing investment in the tools and teams required to pierce through the static and tune in to the collective voice of the customer.