The latest insights from your peers on the latest in Enterprise IT, straight to your inbox.
This article is by Featured Blogger Randy Bean from his LinkedIn page. Republished with the author’s permission.
TD Ameritrade has been garnering considerable public attention these days, much of it due to the high-profile television ad campaign known as “The Green Room.” The ads feature bearded actor Jim Conway calmly talking to investors including the young married couple, the busy mom, and singer Lionel Ritchie about their investment goals. The advertising campaign, which has been running since 2017 and was developed by the firm Havas New York, shows the “approachable Green Room guy” listening without judgment and engaging in discussion about finances and investing goals in “everyday language.”
The ads suggest a focus on each customers’ financial situation and needs, and indeed “hearing the voice of the customer” is a long-term strategic focus for the firm. TD Ameritrade has embarked on a variety of ambitious analytics and AI initiatives in the industry to truly understand the voice of millions of customers, quickly respond to their needs, and tailor highly individualized investment recommendations.
Analyzing the Voice of the Customer at TD Ameritrade
The “voice of the customer” has long been a focus at TD Ameritrade. The company is an ardent follower of its net promoter score, and CEO Tim Hockey has declared that winning on the client experience is the firm’s top priority.
However, the actual voice of the customer has been an elusive entity—for TD Ameritrade and everyone else. Firms can pursue it through surveys and “customer journey” analyses, but it has always been difficult to know what the customer is thinking in detail. Call center conversations are, of course, the direct voice of the customer, but they are inevitably difficult to capture and analyze. The firm employs a large number of call center reps who handle millions of calls per year. Manually tracking the content of the calls is not feasible given the volume of calls handled.
But AI can help. TD Ameritrade has applied artificial intelligence to analyze call center conversation in order to improve the client experience. Beaumont Vance, TD Ameritrade’s Director of AI, Chat and Emerging Technology has led a project to capture, analyze, and interpret a large sample of calls over the last 6 months, and the project is moving from pilot into production.
The first step in the process is to convert speech into text with high accuracy, which is relatively straightforward these days. Then a Natural Language Processing (NLP) model reads through the transcripts, identifies topics mentioned on the call, and analyzes customer sentiment. The model’s analysis is then linked to the customer’s file with the company..
The value of this combined information is that the actual voice of the customer can be linked to the record of activity and behaviors for that customer. For example, it could be determined which web page a customer went to immediately prior to calling the call center. TD Ameritrade can understand this particular behavior by tracking their activity on the TD Ameritrade website (where on the website did they go, what did they do or try to do), what they said on the call, what actions were taken after the call, and what longer-term follow-ups were made. By understanding which web pages might not be meeting customer needs, they can improve those sites quickly and improve the customer experience. This is more efficient both for the customer and for the company.
The goal, of course, is to better serve customers by better understanding what drives their actions and sentiment.. Vance figures that there is 500% more information in the customer calls than in the traditional data upon which most companies rely to perform analytics. Assuming that TD Ameritrade can move beyond the successful pilot and analyze all customer call information, it could have an enormous positive impact understanding customer sentiment and behavior.
As one example of achieving greater customer understanding, the company noticed from analysis of call text that many customers were calling to get information about the gains, losses, and cost basis of their investments. As a result, TD Ameritrade offered a new web page to customers that provides that information with a service called Gainskeeper. Calls and visits to offices for that information declined, and the company could be sure that its investment in the service was valued.
Other AI Work at TD Ameritrade
While the “voice of the customer” analysis is perhaps the most ambitious AI project at TD Ameritrade, there are a variety of other projects and products at the firm that make use of AI. Chatbots are in active use; Vance says that the company employs several popular chat platforms to better support its customers, such as Alexa, Facebook Messenger, Apple Business Chat, WeChat, Twitter and so forth. The goal of these projects is to better serve customers for straightforward issues. Like most such projects in other companies, the chatbot work at TD Ameritrade has delivered considerable value, but will do so even more as the technology matures.
Consistent with the company’s interest in customer relationships, TD Ameritrade employs a tool to aggregate and analyze data about customer journeys. The tool allows understanding of just what customer channels and tasks were necessary for customers to accomplish their goals.
Vance’s last job at TD Ameritrade was overseeing analytics for the retail segment of the company, and he’s a big advocate of automated machine learning as a way to deliver predictive models more efficiently and effectively. That technology is increasingly being used to meet internal needs for machine learning models.
TD Ameritrade is also pursuing a variety of automation options, including robotic process automation. The potential efficiency advantages of these technologies are too great to ignore. However, the goal of this work, like the other projects we’ve described, is not to automate lots of jobs, but rather to improve customer service and free employees to do more creative and unstructured tasks.
Providing the best customer service in the brokerage and personal investing industry is not something that can be done with only human support. Artificial intelligence is necessary to make sense of millions of customer interactions and to help customers make better financial decisions. At TD Ameritrade, AI is making “hearing the voice of the customer” much more than a catchphrase.