Many banks and credit unions spend a big chunk of their marketing budgets on consumer research, including studies, polls, and focus groups. But how useful — really — is the formal research your financial institution pays for? It might bear little resemblance to how people really behave and what they are truly looking for.
Financial marketers also analyze reams of reports that dutifully summarize consumer behavior. Transaction records, going back decades, often are cited as gold mines of consumer data.
It all sounds very promising, but Beaumont Vance says that reality often falls short of the promise.
In fact, Vance, who is the Managing Director of the Analytics Center of Excellence at TD Ameritrade, believes one of the best sources of what consumers are really thinking is right under under your nose. Financial institutions spend tens of thousands of dollars — some millions — on their call centers to compile and record consumer input, and yet this treasure trove of insights never gets analyzed.
Vance knows what he’s talking about. He has had a long and storied career in data analytics and risk management at financial organizations. He was one of three panelists discussing practical applications for data analytics at a banking conference in New York. Vance joined co-panelists Gary Lotts, AVP/Enterprise Data & Insight Platforms, USAA, and Hamilton Faris, VP/Head of Enterprise Data & Analytics, Northwestern Mutual.
What People Really Say And Do
“The disparity between what people say they want and what they go after is stunning.”
“The disparity between what people say they want and what they go after is stunning,” said Vance. As a heavy user of data, he finds this revelation “either depressing or enlightening, depending on how you look at it.”
Unfortunately, said Vance, being rich in data doesn’t mean being wealthy in business intelligence.
“A lot of what we think of as data is just numbers.”
In other words, there’s no story — no value in data without insights.
And even then, some of what marketing executives do to tease “insights” out of their data amounts to the mathematical equivalent of reading tea leaves — statistical malpractice. They attempt to connect the dots of a series of actions by a consumer to draw a conclusion — e.g., “he went here on the website, and then he went there, and then he did that, therefore he’s unhappy about…”
That’s more or less guessing, Vance scoffed.
Read More:
- Banking Execs Must Get Hands Dirty to Dig Up Data-Driven ‘Gold’
- Data Analytics and The Death of The Modern Banking Industry
- Financial Marketers Must Tap Data to Build Digital Cross-Selling Engines
Listening to The ‘Voice Of The Customer’ (Literally)
To learn what really goes on in consumers’ heads requires interacting directly with real, live people. And many banks and credit unions have a mechanism in place to handle such interactions: the call center. However, they almost never use it to their full advantage.
Vance used his own organization as an illustration of the problem. On one hand, many financial companies strive to build websites that can handle the majority of consumers’ needs. Anticipating every combination of needs, and levels of web literacy, can be difficult. So sometimes the call center winds up serving as a human interface to the website.
TD Ameritrade receives 10 million calls to its center annually. This translates to 270 years of recordings piled up every year.
“When a consumer is trying to do something on the website and they can’t find a menu for it, what do they do?” asks Vance. “They call someone to help them do the thing that they were trying to do without calling someone.”
Ideally, after enough website frustrations bubble up to the call center, trends should be spotted and remedies made. Often a recording of every conversation exists. Yet in Vance’s experience, usually nothing is learned from such records.
“Why did they call us?” Vance wonders rhetorically. “We’re in trouble if we can’t even answer that one simple question.”
Sheer volume is the most obvious reason financial marketers aren’t exploiting this rich vein of consumer insight. Vance said that TD Ameritrade receives 10 million calls to its center annually. This translates to 270 years’ worth of recordings piled up every year.
With humans answering those calls, capturing at least some basics about the discussions would seem easy enough. However, that volume — and human nature — keeps that from happening, no matter how diligently you try to institutionalize it. Sure, you can try (as others have before) to mandate that all call center staffers are required to fill out a form on each call. But they never give much thought to those forms, and eventually everyone just gives up on the idea.
“The data we were getting from call center reps was garbage,” gripes Vance.
But lately, TD Ameritrade has begun turning all those calls into data it can act on, using artificial intelligence techniques provided by Clinc. The vendor’s software analyzes the voice calls and identifies patterns and trends.
“AI gives us the ability to do things we couldn’t afford to do with humans,” Vance explains. The software organizes its findings into a database that TD Ameritrade can draw on. Among the benefits of this data has been identification of website details that need to be changed and indications of how the website can be improved.
Given the size of TD Ameritrade’s business, moving the needle just a bit can generate significant results. Vance said that each 1% reduction of call center volume chops costs by $2.8 million.
Vance believes such savings are only the beginning for this new effort.
According to Vance, the database is larger and richer than everything else TD Ameritrade has on its consumer base.
Going forward, Vance says the information will be used to improve such operations as call center management, user experience features, marketing programs, and interactive voice response (IVR) systems — “everything in our ecosystem.”
USAA: Better Serving Those Who Serve
Gary Lotts said USAA has done a great deal of analysis of why people phone its call center. The venerated financial giant has found that there’s actually a kind of “tension” regarding the conversations, depending on which part of USAA is looking at them.
For example, once someone has called in, there is a push and pull between two conflicting goals.
“When a consumer calls, there is a kind of tension between two goals: servicing them quickly to get them on their way, and taking some time to cross-sell.”
The first is servicing the person’s need and getting them on their way quickly. The second is keeping them on the line in order to cross-sell. Asking questions, and reacting to the answers, to determine someone’s needs takes time — the opposite of throughput.
Lotts said a key use of data analytics in his shop is creating positive experiences for consumers. USAA offers banking, insurance, and other financial services to active members and veterans of American armed forces and their families. Thus indications that a service member is being deployed or transferred is a key event that its analytics scans for.
“First, we want to help make that happen, and then we want to make recommendations across our product lines that fit their needs at that time,” explained Lotts.
While this applies person by person, analytics also point up places where USAA can improve offerings across the entire organization. Lotts said that USAA’s greatest challenge has not been in coming up with insights from its analytics, but implementing them.
Because even technology leaders like USAA still possess many legacy systems, turning ideas into delivered products can be difficult.
“Sometimes it’s more expensive to implement the action,” said Lotts, than what it costs to produce the idea based on analytics.
Consumers’ Reluctance To Help Themselves
The ultimate repository of everything that a financial services provider wants to know about consumers it serves, is the consumers themselves. At some level, they know every place they do financial business, their present lifestyles, upcoming life events, and more. Ideally they should be focused on those affairs in some unified fashion. But while this has been the goal of several approaches, notably personal financial management software, not much has produced solid results that satisfy financial executives.
Greg Neufeld, partner at ValueStream Ventures, who moderated the panel, asked how organizations can best serve people when they don’t have the full picture of their finances.
“We’re still struggling with getting people to do anything at all,” said TD Ameritrade’s Vance.
He pointed to the hopes of sites like Mint that were meant as places where consumers could record all their financial data for a unified view. But that approach has been stymied by the growing reluctance of banks, credit unions, brokers, and other financial providers to permit screen scraping.
There’s also consumers’ typical resistance to periodically making manual updates to such central data repositories. Many are allergic to keyboarding.
“I don’t foresee it happening in my lifetime,” said Vance, “unless the federal government decided to mandate that everybody centralizes their financial data in one place.”
Northwestern Mutual’s Hamilton Faris said his firm has had some success collating more data from consumers than in the past.
“But it’s a slog,” he said. “It’s not easy.”
At the same time, many consumers tend to not read much that would help background them on their financial choices — even from their own financial institutions.
In fact, Neufeld noted that while providers often set up advisory emails and other messages for consumers, often these personalized messages go unread. He pointed out that even checking on their balances can be a pain point for some consumers.
“You have to train them to care about their finances,” suggested Neufeld.