Merrill Seeks an Edge In Interactive Services
March 16, 2012
Merrill Edge provides investment products and services for the "mass affluent" customer, defined by Bank of America as people with $50,000-$250,000 in total household investable assets.
Merrill Edge reaches those customers through two channels: an online site that serves 550,000 "self-directed" investors, and the Merrill Lynch Advisory Center, which offers "team-based" advice and guidance.
Before Bank of America acquired Merrill Lynch at the end of 2008, the advisory center was 100% "center-based,'' said Wendy Delp, director of client services and operations for Merrill Edge.
Centers at that time meant fully licensed financial advisers operated out of national call centers. Currently, there are three: one each in Hopewell, N.J., Jacksonville, Fla., and Chandler, Ariz. About 750 financial advisors operate out of them. Now, an additional 600 advisors operate out of Bank of America banking centers.
Delp's group supports both self-directed clients and advisory center clients.
To improve service, the group has launched two projects since the middle of last year, Delp told attendees of the NICSA 2012 conference in Miami in February. These include a "co-browsing" initiative, designed to help customers with their online experiences, and, a speech analytics program that is invisible to the customer, but helps identify problems in a systematic way so they can be resolved.
This program was put in place in the middle of 2011 as a feature of MyMerrill, a wealth management site. But clients who access merrilledge.com can obtain the service from technical support associates who serve both Merrill Edge and MyMerrill.
With co-browsing, customers get live assistance from technical advisors who can browse the site along with them, resolve issues and point them in the right direction.
The effort is useful for "some of our clients who are less tech-savvy,'' said Delp. "They want to get out and see their account online. They want to place trades. But they have trouble enrolling their account or just finding things on the web site.'''
"Co-browse allows us to see what the client sees on their computer. It's basically a page-share application. If they click a button or go to a website,'' she says, "it allows us to see their computer. That has helped us tremendously, particularly with our older clients who are still using dialup and believe it or not there are still some of them out there."
What co-browsing does is allow Merrill to get problems resolved faster ... and more cheaply.
"Some of these tech support calls were taking upward of 30 to 45 minutes,'' Delp said. "At $4 a minute, that just got too expensive for us to continue to manage.''
After all, at that rate, one half-hour call was costing the company $120.
"The co-browse has allowed us to point to things on the computer for the client and get them through the online process, faster,'' she said.
Improving the customer experience and reducing costs, at the same time.
By their nature, call centers build up huge amounts of data, through all the conversations that take place with customers.
But those mounds of data are not typically in a format that can be usefully managed, interpreted, analyzed and acted on. Because it's all spoken.
Enter speech analytics. This is a system for converting all those spoken words into text, then scanning the results for key words that can indicate problems experienced by a wide range of customers, how well marketing efforts are performing and how broadly a product is getting adopted (or not).
Introduced late last year, the speech analytics program was launched with software developed by an Israeli technology firm, NICE Systems.
The software is integrated into Merrill Edge's call reporting system at its call centers.
Managers build a lexicon of words that helps them understand the client experience. These get organized into different categories, so that, for instance, the company can see if problems emerge, for instance, in how its interactive voice response system is operating. These are the automated systems for handling client inquiries or steering them to the right human being for a live response.
For a category such as "client dissatisfaction,'' key words or phrases that get tracked might be: "You people," "ridiculous,'' "frustrated," "hate this," and, naturally, "want to speak to a supervisor."
The system will find all of the calls that contain words like this, from all conversations with customers. Then it will assign a confidence level, to try and separate the real problems from minor ones.
The confidence rating increases when more and more of the key words show up in a conversation or the same key phrase, keeps getting repeated.
"So if you go in and look at a call with a 90% confidence level, you can be pretty sure that that client is not happy," Delp said. "So you want to target what is going on in those calls.''