Published: 1Q, 2009

Do you know who your customers are?

Better find out.

In 1983, a focus group of about 50 people all said that online banking’s time had come.

That year, Chemical Bank introduced an online banking product called Pronto.  Using a computer modem, phone line and an Atari computer, bank customers could conduct business online at their home or office.  Chemical Bank spent $100 million promoting Pronto, which met crashing indifference from the banking public and caused Chemical Bank – later to be absorbed by Manufacturers Hanover – to take a $10 million writedown on the product.  What went wrong?

The easy answer is that the technology – and the population at large – just weren’t ready.  The less-easy answer is that Chemical Bank misinterpreted the research.  In order to determine if its online banking would work, Chemical Bank developed some focus groups.  Virtually all of the people recruited in one key group were bankers; the majority were early-adopters of computer technology.  In other words, they were NOT the broader base of customers Chemical hoped to reach.  That key focus group expressed huge enthusiasm for online banking.  Chemical had never conducted a random probability survey of all customers to find out if they wanted home computer-based banking as well as the key focus group participants.  Unfortunately, this typs of error is even easier today with online surveys, panels, and the mistaken belief that anybody can ask good survey questions.

Today, Internet based survey sites such as Survey Monkey and Constant Contact make it appear as if surveying has gotten easier, says Larry Cohen, director of Consumer Financial Decisions group at SRI Consulting Business Intelligence, based in Princeton, NJ.  But, in fact doing a good survey has gotten even harder.  These services can collect and disseminate information faster and more efficiently than ever before.  But the burden of selecting the right people, i.e. the sample – customers or potential customers – to survey, and asking the right questions about their banking relationships, still depends on the researcher.  Which gets back to the ‘know your customer’ ethos. If you don’t collect the right information, the right way, how can you know if you’ve come up with the right answers? And, any decision made on this data runs the very real risk of being another Pronto decision.

Once armed with the right data, banks have to execute properly, following up with good old-fashioned customer service, making the marketing effort a twofold endeavor.

Life-cycle logic

Thirty-five years after Chemical Bank’s Pronto venture (which became a Harvard Case Study), online communication has come of age. But, ironically, it can make banking outreach to customers more difficult rather than less.

Marketing strategies that take a predictive stance based on life cycle can come out ahead of the game.

John Matheny, director of sales and marketing at Austin TX-based bank consulting firm Brintech notes that as individuals or households accumulate more money, or evolve into a situation in which they have more disposable income, they use more bank products. Banks should start with a basic understanding of the market segments within their market area, and alignment of products and services to fit those segments’ usage trends, he says.

Historically, younger households, especially those with middle incomes tend to use checking with debit and ATM cards, overdraft protection, credit cards, and mortgages more than those in other life stages. Once the nest is empty and people are in pre-retirement, they focus on accumulating as much wealth as possible, i.e. the “sprint to retirement.” Once in retirement they live off assets and income. Because of the impact of greater life expectancy, delayed child rearing, many years of low or no savings, and aggravated now by the economic meltdown, Cohen notes the emergence of “revolving retirement” life stage – intermittent work, or work at a lower level of intensity. This interim stage between pre-retirement and before traditional retirement, could help consumers make their savings last longer, delaying the need to draw on Social Security, which results in bigger benefit payments in the future, and decrease the uncertainties of outlasting ones assets. This phased retirement was gaining in popularity even before the financial upheavals of late 2008, but now, except for the very wealthy, it is becoming a necessity, Cohen says.

Gathering data

Matheny notes that the first contact at the new accounts desk is where the process of identifying a customer’s circumstances begins. The identification process continues through each and every contact with the customer as the relationship develops.

A key customer model for banks is the “total share-of-wallet” – how much of their customers’ total assets are with the bank? “You may know their income if they’re a borrower, and you may have some notion of assets, because you are managing some of them,” says Tony Coretto, managing partner of PNT Marketing Services Inc., based in Long Island. “But you won’t know the customer’s total wallet or your total share-of-wallet,” he says.

Primary demographic information comes from the customer – age and sex. Social Security numbers, while they are keys to pull in additional demographic data elements, must be stored and used with the utmost caution due to privacy and security concerns. To dig deeper, banks must purchase secondary demographic data from sources such as the U.S. Census or any one of a handful of multinational companies that track data using proprietary methods combined with data-sharing information provided by subscribers.

This data does not include individual households, though. The data is shown in an aggregate, the way Census data or employment data is drawn. To make the information more specific, data subscribers may, for example, go by ZIP code and age to overlay the data on their market segments.

Finding the most profitable segments

A bank in an affluent region, for example, might have three large customer tranches – affluent retirees 65 and older, no children in the household, $1 million home and $1 million in investable assets. That’s 20 percent of its customer base. The next tranche, aged 45-64, own homes worth $500,000 and have $500,000 of investable assets. The remaining 20 percent are in micro-segments, based on banking relationship, their age, and whether they own or rent.

This primary research is descriptive, but it does not reveal the customer attitude toward the bank. To find out, banks can create several different surveys to determine where the difference lies in banking attitudes – how the coveted customer prefers to bank, whether it’s at a branch, on the phone, or online, for instance.

With profitable customers, retention is key, Coretto says, noting that a good retention strategy may offer premium rates and a tailored bundle of products and consolidation strategies.

The most profitable 20 percent of customers 65 and older may utilize personal bankers, and they may prefer direct mail to e-mail because they might not have e-mail.

Another group, still in the wealth-growing stage, might be good prospects for a HELOC to pay for college in a tax-deductible way.  This group communicates by e-mail, not direct mail – they throw that out, Coretto says.  At a branch, they’re best served by a private banker and/or a kiosk.

The remaining 20 percent is more problematic.  “You just don’t have enough discrete data elements to drill down to the very last level” of that last 20 percent, he says.  That last tranche might break out into 12 different socioeconomic groups which are not different enough to warrant different levels of service for different levels of profitability.

“It is virtually impossible to broadly market to the exceptions,” Brintech’s Matheny says, “but it is relatively easy to identify the exceptions within a customer base and build relationships and increase retention based on assisting with those circumstances.

Great idea, but does it pay?

Pittsburgh, PA-based Quest Analytics uses statistical methods to identify groups with similar characteristics.  “When you analyze the data, you gain a much better understanding of your customer.  By understanding customer behavior, you can execute a personalized strategy to help each individual customer” says President Karl Keller.  Profit levels are higher and it is easier to retain customers in uncertain times.  “If you don’t do the analysis, you don’t know how they will use the product which could lead to cannibalizing your other products.  Measurement is crucial -- if you don’t measure it, it probably didn’t happen.”

Offering a limited time only higher CD rate to a targeted age demographic has its risks.  “That strategy backfires if that customer isn’t profitable for you,” says PNT Marketing’s Tony Coretto.  “When you do your profiling and look at profitability by segment, the person who fits the ideal demographic may not be the most profitable for you if they’re single-product customers who are rate shoppers, for instance.”

As potential sources of loan funding, the cost of the deposits of standalone CD customers must be considered in offsetting returns on personal loans.  It’s too simple to say all deposit customers are losing money and all loan customers are making money, Coretto advises.  “Your financial model has to be sophisticated enough to match deposits to loans in the portfolio and come up with the appropriate transfer pricing.  That way these absolutely necessary CD customers don’t look like as big a drag on profitability, because those deposits are funding someone else’s loans elsewhere in the portfolio.”

Capitalizing on convenience

The economic turmoil of the last year has created many new scenarios for banks’ customers and can provide a vital opportunity for banks to increase the services they provide.  The key for banks is recognizing the value customers see in them and capitalizing on the convenience, and simplicity they are looking for in these trying times.

With retirement plan balances and home equity both shrinking, far fewer people will voluntarily retire in the near future.  Of those ‘forced’ to retire, many will immediately start looking for other work, even part-time.  This situation is creating a significant change in how consumers behave, with long term ramifications for financial institutions and advisors.  In Cohen’s view it is, “the ‘perfect storm’ of society experiencing three tipping points simultaneously: Baby Boomers in transition and redefining retirement, the impact of financial convergence, and a consumer shift from conspicuous consumption to conspicuous conservation. “All of these will reorient consumers’ priorities.  The key to meeting consumer’s needs in these uncertain times are trust, simplicity and convenience,” he says.

Christina P. O’Neill is Custom Publications Editor for The Warren Group, publisher of Banking New York

 

Outreach 101

Once you select the proper customers through the correct demographic research (see sidebar, page 12), how do you talk to them?

On boarding – customer interaction during the first 90 to 180 days after a customer opens an account -- is the Welcome Wagon of bank marketing. The first 30 days is critical, says Tony Coretto, managing partner of PNT Marketing Services Inc. “Introduce them to all your services, remind them you are happy to have them as a customer, make sure they’ve gotten their checks, basically check in with them and make sure they understand all the services you offer.”

An outreach might start, for example, with a CD offer to help consolidate a customer’s balances. Then, enrolling the customer with a debit card, bill payment and direct deposit, makes the bank-customer relationship much “stickier.” Once these “sticky” behaviors have been established, the customer is more likely to stay – and more receptive to cross-selling other products such as HELOCs, student loans, mortgages and reverse mortgages, which will vary by market segment.

After the first 90 days to 18 months, the banking relationship should encompass the customer’s entire lifecycle. By then, a bank should know its customers’ attitudes toward the bank based on customer segment, and it can tailor its communication depending on what the segment is.

Karl Keller, president of Quest Analytics, a bank consultant based in Pittsburgh, PA, says banks do a good job of servicing and deepening relationships with customers who visit branch offices regularly. However, with the growth of electronic banking and services such as remote deposit capture, he says, “You’ve now given your customer 20 different reasons to NOT visit your branch.” The real question is: What is your bank strategy to deepen relationships with your customers when you don’t see them as often? Without a consistent program to stay in constant contact with your customer, you risk becoming a commodity and always having to compete primarily on price.”

Quest Analytics helps banks use demographic and transactional data to map changes in custome behavior. An example: a customer whose deposit levels are almost enough to qualify for a higher money market interest tier. In most cases, the customer did not know he/she was getting close to the next interest rate tier. A courtesy call is made to advise the customer of the situation and is gently encouraged to add the extra cash. This is only done after they’re asked the right questions to make the best recommendations to meet their financial goals.

Keller says he often meets resistance from bank CFOs to such customer contact, due to concern that the customer might be inspired to rate-shop elsewhere. To the contrary, he says, a proactive call strategy, which includes an assessment of customer needs, more often leads to customers taking deposits out of competitors’ banks.

"The bank’s approach should be that selling is helping,” Keller says. “No one likes to be sold to, but everyone likes to buy something.”

 

Customer surveys – their past and future

The ease and inexpensive nature of online surveys and panels can conceal how easy it is to make basic mistakes in selecting survey respondents and panelists. Larry Cohen, director of Consumer Financial Decisions group at SRI Consulting Business Intelligence, based in Princeton, NJ, cites some sample selection basics that transcend the electronic age. The number of respondents chosen to represent a wider population does not need to be huge -- it is much more important to choose the right sample of respondents, and randomness is the critical key.

Every potential respondent should have a known chance of being included in the sample in order to ensure the randomness of the sample. Perhaps the most famous example of where this was not done was the infamous election poll in the 1948 Dewey-Truman presidential race. A poll of magazine readers had Dewey as the clear front runner. However, the pollsters did not take into account that most of the readers of the magazines selected for sampling were much more likely to be Republican. More than three decades later, Chemical Bank did something very similar when it selected only computer-literate banking professionals as its core panel (see main story).

In order to save the time and money it takes to recruit a random sample for every survey, firms pre-build panels of respondents and use the same people repeatedly. Some of these panels members drop out with every survey they do and have to be replaced, Cohen says. But in order try and retain the original representative nature of the panel they cannot be replaced randomly – they have to be replaced with people with similar characteristics of the ones who drop out. Since certain types of people tend to drop out of panels more quickly than others (and are less likely to be on a panel to begin with) the panel quickly loses its randomness and its ability to project to the total population.

Panels hold even more dangers for financial studies. Panel participation requires frequent repeated interviews. For some areas – particularly financial studies -- panel participants subconsciously or even consciously start to pay more attention to ads, articles, their own experiences, stories and what others say about financial services. They become more knowledgeable about these topics and ‘lose’ their representativeness, in this case to be relatively uninformed about financial matters, and they become financial survey respondent professionals. Results based on these types of samples will mislead decision-makers into believing consumers are more knowledgeable and interested in financial services than the real population. As any panel progresses from its initiation, it becomes less and less representative, Cohen says.