Reader's choice: The quest for customer focus

Reader's choice: The quest for customer focus

Getting closer to customers isn't only about building an information technology system. It's a learning journey of four stages, each with its own obstacles, and each requiring people and units to coordinate in ever more sophisticated ways

More and more CEOs are hoping a stronger customer focus will be the antidote to escalating commoditisation pressures. But as the frustrations of myriad companies can attest, getting closer to customers is not just a matter of installing a better CRM system or of finding a more effective way to measure and increase customer satisfaction levels. Tools and technology are important. But they're not enough. That's because getting close to customers is not so much a problem the IT or marketing department needs to solve as a journey that the whole organisation needs to make. The companies that do it well follow a surprisingly similar path, passing the same milestones and, in many cases, struggling with the same problems. The journey can be arduous, it takes a long time - years, not months - but there are rewards all along the way. And for those organisations that have gone the distance, the payoff is remarkable.

For Continental Airlines, the journey began when the company was emerging from bankruptcy and needed to know more about the profitability of its individual customers. One of the first things it uncovered was a service mess that was costing the airline millions of dollars every year.

Continental took a systematic look at how passengers were treated when a plane was significantly delayed, when they were bumped from a flight, or during some other unfortunate event. What it found was that compensation was offered on an arbitrary basis by the gate agent, and, somehow, the lowest value customers were, on average, receiving the highest compensation.

Worse, some passengers were finding ways to be doubly compensated; a customer who was bumped from a flight might first approach a gate agent, pick up a voucher for a free flight, and then minutes later telephone the airline and ask for another. The representative answering the phone would have no way of knowing that the same request had just been filled.

It was only when the company began to look at customer information in a more holistic fashion - gathering, consolidating, and analysing all of its customer interaction information in a single pool - that it was able to correct such inefficiencies.

Now everyone who is delayed for, say, nine hours gets the same compensation, and when a gate agent hands a passenger a flight voucher, that transaction is reflected immediately in the customer information database. The passenger will be denied a second voucher even if he gets to a phone within a few seconds.

For Royal Bank of Canada (RBC), the quest for customer focus began when the company discovered it knew much less about the needs of its customers than it thought. RBC is Canada's largest financial institution, with more than 12 million personal, business, and public sector clients and offices in some 30 countries worldwide. In 1996, like most financial institutions at the time, RBC had been investing heavily in making banking as convenient as possible, on the assumption this would attract new customers and increase loyalty. It extended banking hours. It built new branches and installed more ATMs.

It added online access. It created insurance, investment, and other new services. But to the company's surprise, a survey of more than 2000 current and potential customers revealed people didn't choose a bank on the basis of how convenient it was. RBC scored very well on that measure. But, as the survey clearly showed, that was merely table stakes. Instead, what customers wanted was a bank that demonstrably cared about them, valued their business, and recognised them as the same individuals no matter what part of the bank they did business with.

Based on this insight, RBC set a goal to systematically manage all of its customers at every one of the millions of points at which they came in contact with the bank - a prospect that was daunting, to say the least. To its credit, over the last nine years RBC has learned how to reorient the focus of its entire organisation away from products and distribution and toward the real needs of its customers.

The results are telling. Dividends swelled from 68 cents per share (in Canadian dollars) in 1996 to $1.72 per share in 2003, driven by a 20 per cent increase in high-value customers and a 13 per cent rise in average customer profitability between 1997 and 2001. Between 2000 and 2004, the percentage of customers that purchased the bank's high-margin packages of bundled products and services doubled, from 35 per cent to 70 per cent, and the success rate of sales leads driven from promotion events rose to 45 per cent.

(Compare this against the 2 to 5 per cent response rate typical of standard marketing programs.) Along the way, RBC has won a host of information technology awards for its innovative customer-facing computer systems.

The Continentals and RBCs of this world are as exceptional as their exceptional results. But they are not unique. What are they, and others like them, doing right? To answer that question, we spent two years conducting an in-depth study of 17 companies that have made substantial progress toward becoming more customer focused.

This diverse set of businesses ranges from financial institutions like RBC, to the gaming giant Harrah's Entertainment, to the massive telecom carrier SBC Communications. What we found, at a high level, was that customer-focused companies consistently embrace three concepts.

First, they know they can become customer focused only if they learn everything there is to learn about their customers at the most granular level, creating a comprehensive picture of each customer's needs - past, present, and future. Second, they know that this picture is useless if employees can't or won't share what they learn about customers, either because it's inconvenient or because it doesn't serve their interests. Finally, they use this insight to guide not only their product and service decisions but their basic strategy and organisational structure as well.

Over time, these companies enable and enforce coordination between internal units at progressively more sophisticated levels, they find new ways to manage the flow of information, they develop routines for decision making that incorporate customer preferences, and, ultimately, they shift the locus of their customer-focused efforts away from a centralised hub to a more disbursed set of activities that spans the entire enterprise.

Each company we have studied has followed a strikingly similar path in its journey, a path that runs through four distinct stages. Skipping stages might appear to speed up the process, but in the end it denies the organisation the sure foundation it needs to build a lasting customer-focused mind-set.

Stage one: Communal coordination

The journey begins with the creation of a centralised repository of customer information, which records each interaction a customer has with the company. Creating this repository is a two-step process.

First, organisations bring together and standardise information drawn from customer touch points throughout the firm into a single pool. Second, they organise this information by customer; that is, they make the customer - rather than the account, the purchase, the product, or the location - the fundamental unit of analysis.

The definition of a customer may not always be obvious. For Continental Airlines, for instance, customers could be defined as travel agents, corporations, or consumers. As a result, organisations may have to manage and collate their interactions with several interrelated customers together.

Gathering, standardising, and organising customer information that comes from all across the organisation requires companies to establish a coordination infrastructure. The amount of coordination called for in this stage can be substantial, but it is not necessarily complicated; we call it communal coordination.

Organisational units need not contact one another directly. Instead, each group contributes its information to the communal pool separately from the others and then taps into it as needed. In the companies we studied, a neutral entity like IT typically controls and oversees the pooling process. That's for two reasons. First, employees of a neutral entity like IT have the technical skills to normalise and cleanse information as it comes into the common repository.

Second, and more important, such staff tends to be free of operational biases. Unlike sales, marketing, and other groups that create and use customer information, neutral entities like IT don't concern themselves with the actual value of the information; they care only that it is accurate, clean, and easily accessible. However, ownership of this process requires employees of the neutral entity to possess a unique mix of skills - an understanding of both the technology and the business needs of all the different groups that rely on customer information to make key business decisions. Such talent is not commonly found in most organisations.

The concept of a communal information source is relatively simple, but it requires a substantial investment in both time and technology to make it useful. There is, of course, the challenge of overcoming political boundaries, as people often resist sharing information - and resist losing control over it even more. But at this stage of the journey, it's simply the sheer volume of customer information that tends to make the pooling process so long.

At Continental Airlines, it took more than four years. Just for a start, it took six to nine months to properly clean and aggregate the information as it came in from the finance, marketing, operations, and other units.

The need to capture information at a granular level is another reason the process is so time-consuming. Continental's database includes as fine a level of detail as a customer's choice of seats and preferred methods of booking; the number of times his flight departed on time, was delayed, or was canceled; and any time his luggage was lost.

If the task is onerous, the payoff is high: The goal of collecting information at so comprehensive a level is that Continental no longer needs to know in advance what business questions it might wish to ask. The repository has within it the potential to answer just about any question.

Pooling the data took even longer at Harrah's Entertainment: Six years. Unlike Continental, which was taking over a set of processes that had been previously outsourced, Harrah's had to overhaul internal customer information management systems already in place in numerous properties scattered across the United States. When Harrah's began this process in 1991, competition in the gaming industry was local.

Casinos in Las Vegas competed with rivals along the Strip, making ever more costly upgrades to woo customers to their tables from the tables next door. But initial research, which the pooled data confirmed, indicated customers weren't really choosing among competitors on that basis. What they really wanted was to be properly recognised and rewarded when they visited Harrah's in another market. So Harrah's sought to shift its strategy away from focusing on competitors toward standardising and improving its customer experience in all of the company's properties, thereby creating a national brand.

Creating a single view of customers is valuable in its own right: It generates opportunities for cross selling, it reveals glaring errors in customer service, and it can point the way to efficiencies that reduce costs. But more important still, consolidation sets the stage for the next steps on the journey toward customer focus - in two important ways.

First, as individual business or functional units are forced to share information, companies begin to see a shift in mind-set. Employees in one business unit learn to recognise that "their" customers are shared assets, valuable to other units as well. This limited amount of coordination lays the groundwork for much higher levels of coordination to come.

Second, the central repository of customer information itself serves as a building block for the next stages. Continental Airlines in the mid-1990s had 35 to 40 domestic databases and another 50 international databases; more than half of them were dedicated to customer information, but nearly all of them housed some important customer data. Now it has two databases, one for customer analytics and modelling, and one for operational data.

Stage two: Serial coordination

In stage two, companies go beyond just assembling customer information to drawing inferences from it. Coordinating gets a little trickier as the centralised coordination role expands to manage not only the continued collation of data but also a sequence of tasks performed by certain functional units so that information can be analysed and the resulting insights shared throughout the company. Wecall such a coordination architecture serial coordination. The sequence typically starts when the collated information from stage one passes to business analytics experts (who frequently reside either in marketing or in a separate unit of their own). They analyse the information and then pass along their results to users in the business units, who identify how best to apply it in marketing efforts, building on their knowledge of the local markets. The handoff from one unit to another may not occur spontaneously; one of the organisational units may need to take a leadership role to ensure that the entire sequence of steps is accomplished and properly coordinated.

In some organisations, the information is employed for more than sales and marketing activities, to analyse and improve a broad range of enterprise operations. For example, Continental was able to mine its customer information to configure its flight network more efficiently. Previously, the airline could analyse only the profitability over time of each route separately by tracking the number of passengers on each flight and the average fare they paid. It did not know whether those passengers were coming or going or what routes they'd previously flown. Now, with the data pooled from all organisational units, the route optimisation teams can examine the total revenue generated by each passenger and the pattern of that individual's travel. Viewed in that context, a flight that was previously considered unprofitable - and, consequently, a target for elimination - might turn out to be a key connection between airports used by a substantial pool of very profitable customers.

Learning to use the communal data in this way requires substantial coordination. Employees in Continental's planning and scheduling group managed the overall process to ensure that the serial coordination actually occurred. In this instance, as they began their analysis, the members of the group realised they needed input from several different areas of the business, including operations, pricing, sales, marketing and finance. They approached each one in turn for its input into the analysis of the flight network. The operations group provided information about how scheduling changes would affect when and in what way airplanes were serviced. Pricing then added information on how changes in ticket prices might influence customers' choices. Sales subsequently assessed what channels would best suit the offering. Marketing got involved to plan the launch of new flight segments. And, finally, finance made sure the assets of the organisation were being put to the best use.

The Royal Bank of Canada saw similarly powerful benefits from analysing its pooled data. In one instance, RBC was examining the effectiveness of one particular service offering. The package combined a checking account, credit card, and some other services, like the ability to pay bills at its ATMs. It was popular, but RBC's analytics team found that nearly 60 per cent of the time these packages were unprofitable. In the past, that discovery might have forced the bank to discontinue a product that customers liked or to raise fees for the package, either of which might have solved the bank's problem but not the customers'. With its transaction-level data, however, the company could pinpoint the source of the problem: The ATM bill-paying process wasn't automated. Bank employees had to take the paperwork from the envelopes inserted into the ATM and enter the transactions in by hand.

Serial coordination was overseen by the marketing and strategy group, which initiated the project. The analytics group provided the analysis, the product management group re-evaluated the content and pricing of the product bundle, and the finance group studied the impact of the various options on the company's performance.<p/>In the end, after considering the input from all of these groups, the bank kept fees the same but added low-cost telephone and online bill-paying services to the original package. Happy customers began using these convenient options of their own volition. And a year later, 90 per cent of these packages were profitable.

Serial coordination is not spontaneous and is fraught with obstacles. Traditional roles and structures create natural barriers to spreading information and lessons learned throughout the company. Some changes to a company's social and organisational structure will be required to overcome them. One of the most significant barriers can be a lack of trust between the group that collates the information, the analytics experts who manage it, and those who apply it in the line organisations.

The best way to whittle down those barriers and build trust is to show some early successes. For example, when RBC began using customer information to target sales activities more precisely, the central analytics team began creating much shorter lists of customers for bankers in the individual branches to contact with new offers. The bankers were initially - and understandably - sceptical of the effort when they started getting 20 names instead of the 300 they were accustomed to, and they hesitated to use the pared-down lists. But they quickly recognised that the new lists yielded much better response rates. That gave the bankers far more confidence in the analytics team.

The second stage in the quest for customer focus usually uncovers critical gaps in employees' skills. Most people are unaccustomed to having so much customer information to work with. Often, those with statistical skills lack business savvy; those with operational knowledge are not comfortable analysing data. It's difficult to find people who can be "bilingual".

In general, we found that best-practice companies centralise this analytic capability because it's not possible or practical to hire people with PhDs in statistics for every unit of the company. With the entry of COO Gary Loveman and his new leadership team in 1998, for instance, Harrah's created a central marketing-analytics team tasked with interacting with the various casinos and ensuring that all of them used the customer data effectively. This arrangement was not something that everyone in the corporate organisation or the individual properties was good at or comfortable with, which led to some turnover in both units.

Stage three: Symbiotic coordination

Stage three is a step jump in terms of complexity and the need for coordination because it requires that companies shift their focus from an analysis of past customer interactions to anticipating, and even shaping, the future. They begin to ask questions like: Which customers will be likely to switch to a competitor? Which are most likely to buy a new product or service in the future? Which are most likely to pose an unacceptable credit risk? Addressing these questions requires organisations to move away from the one-way information flow that characterised the previous stage toward a dynamic give-and-take.

We call this symbiotic coordination: Information and decisions flow back and forth between central analytics units, operating units, and marketing, sales, and other organisational units - and even laterally among the organisational units themselves.

In this stage, companies embrace an experimental process comprising four discrete sets of activities: Creating models to predict customer behaviour; experimenting with various interventions designed to alter customer behaviour; measuring the results of these interventions; and using feedback from the front line to improve the models and subsequent campaigns. In true scientific fashion, companies often set aside a control group and compare the activities of those customers who received an intervention with those who did not.

By repeatedly altering the experiments and carefully measuring the results, companies learn over time which alternatives have the greatest impact on customer behaviour.

For example, SBC wanted to decrease the number of customers that might defect to the competition. So using the analysis done in stage two of virtually every interaction between the company and its millions of customers, SBC created various defection models to predict the likelihood that an individual would switch to another telecom carrier. It then developed and experimented with various marketing interventions designed to hold on to those at-risk customers. In one instance, SBC learned that people who subscribed to the company's SBC Yahoo! DSL service were significantly less likely to switch their local phone service from SBC to another vendor.

Using propensity-to-buy models generated from its information pool, SBC then identified customers with the greatest likelihood of purchasing DSL. This in turn allowed product managers to identify and target only those individuals who would be profitable in a 12-month period. In one campaign alone, SBC was able to reach only its most profitable potential DSL subscribers, without significant marketing expense.

Symbiotic coordination requires people in several units who have no formal reporting relationship to interact in spontaneous and unsystematic ways through a constant give-and-take. Work is not handed off serially from one group to another; people are learning together in real time. Pulling this off typically calls for some major structural changes.

Most companies take one of two approaches to creating the needed links: They reorganise the entire company by customer segments that cut across product, technology, and geographic boundaries. Or they add new organisational units whose job is to ensure coordination between the centralised IT and analytics experts and the front line. Either way, it's a big job.

Royal Bank of Canada took the former approach. The bank was previously structured around products; employees attended to "mortgage" customers or "deposit" customers rather than to "RBC" customers. Now the company is structured around three customer segments: Premium, standard, and foundation customers, each of which cuts across all of the product lines.

To preserve a degree of accountability for sales, RBC also created a matrix structure, laying product segments over the customer segments, rather than obliterating the product segments altogether. As expected, the matrix structure engendered some tension between the customer and product organisations; employees focused on products are interested in selling their own products; employees focused on customers are rewarded for maximising the value of all customers to the organisation as a whole.

To avoid confusion, RBC's leaders have made it explicit that customer segment employees have the final say in all of the product-oriented staff's customer-related budget decisions. It takes patience and time to make such dramatic changes. RBC Banking vice chairman Jim Rager waited nearly five years after beginning the customer focus journey before undertaking this reorganisation. To make the change more palatable, he wanted employees first to see some early successes with the new customer-focused approach.

Stage four: Integral coordination

If in the symbiotic stage, companies shift their focus from the past to the future, in the integral coordination stage, they focus on bringing a now-sophisticated understanding of their customers into the present, incorporating that understanding into all of their day-to-day operations. Companies start to move past discrete, formal initiatives to weave customer focus into the informal values and daily behaviour of all employees. Customer focus begins to define the organisation and pervade its every aspect.

Whereas previously, central marketing, IT and analytics groups were the primary drivers of customer-focused initiatives, now activities are extended down into the line organisation, where employees are given the autonomy and latitude they need to focus on the customer in virtually every action. Continental, for instance, allows nearly all employees in the company access to its customer information - and it also provides them with access to the experts who can help them analyse and use it. Technology director Anne-Marie Reynolds observes nearly half of her department's time is spent helping employees access and understand customer information.

At this stage, companies are coordinating key activities across vertical and horizontal boundaries, which are in many cases irrelevant to customers. We call this integral coordination. Companies can build informal overlays that transcend organisational boundaries, bringing together people with a passion for some particular customer-focused activity into centres of excellence.

At RBC, a data warehouse steering committee, with representatives from all of the bank's lines of business - RBC Investments, Insurance, RBC Banking, and Commercial Banking - establishes priorities and the order in which customer information projects will be funded, making sure that these projects are aligned with overall strategic objectives and that existing corporate roles and structures do not prohibit learning transfer.

When customer focus becomes institutionalised in this way, technology can not only support but even automate decisions. At Harrah's, for instance, technology helps employees to more productively allocate its scarcest resource: Its hotel rooms, which run at 95 per cent occupancy year-round. Customers spend more when they stay at the hotel than when they just visit the casino, so Harrah's wants to be able to put customers with the highest potential into the rooms.

A new system that can match up customer profitability measures with occupancy predictions helps employees book rooms in a way that dynamically optimises profitability. Very profitable customers could be given rooms for free, while unprofitable customers could be charged the highest rate. The results of this and other customer-focused programs have been outstanding; profit per available room increased 30 per cent between 1999 and 2003, which equates to more than $20 million a year added to the bottom line despite a significant increase in costs, as the number of rooms in the network has expanded.

While companies can and should distinguish between more and less desirable customers, they should not forget that lower-value customers may over time become more profitable. Indeed, very few of the companies we studied tried to get rid of lower-value customers. Rather, we observed, they sought to understand such customer segments, to identify profitable characteristics in seemingly less-profitable customers.

For instance, Harrah's Total Rewards program recognises people based on their annual value so that lower-but-steady spenders can aspire to perks that high rollers enjoy, such as shorter lines at restaurants. Harrah's then modified Total Rewards in 2003 to allow customers to carry over points from year to year so that they were treated according to their true long-term value.

The challenges in this stage are monumental, primarily because integral coordination requires a major shift in attitude on the part of so many employees.

Shifts in attitude cannot be forced. Employees can only be nudged, pressured, coaxed - and provided incentives. Harrah's invested US$40 million in 2004 alone to reward those employees who according to their managers had delivered outstanding customer service. And after changing its incentive structure and providing comprehensive training, Harrah's showed how seriously it took its customer focus initiative by eliminating the recalcitrant resisters.

As a result of these efforts, Harrah's increased the share of its customers' gaming wallet from 36 per cent in 1998 to 43 per cent in 2003. SBC and Royal Bank of Canada have also had to change their incentive structures so that they not only reward sales but also encourage cross-unit cooperation and client-focused behaviour.

Companies have poured an enormous amount of money into customer relationship management, but in many cases the investment hasn't really paid off. That's because getting closer to customers isn't only about building an information technology system. It's a learning journey - one that unfolds over four stages, each with its own obstacles, and each requiring people and units to coordinate in ever more sophisticated ways. Companies that recognise this will invest their customer relationship dollars much more wisely - and will see their customer-focusing efforts pay off on the bottom line.

Harvard Business Review

Join the CIO New Zealand group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.

Join the newsletter!


Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.
Show Comments