benefits/costs of using within-firm secondary data

Business Insight (A Special Report); Innovation: Staring You in the Face; The path to new products might start with the customer data you've already collected; You just don't realize it
Ranjay Gulati, James Oldroyd, Phanish Puranam. Wall Street Journal. (Eastern edition). New York, N.Y.: Sep 22, 2008. pg. R.9
Abstract (Summary)
At Royal Bank of Canada, the decision to look again at rejected loan applications was a dramatic switch for the risk- management department, says Cathy S. Burrows, director of client strategy at RBC Insurance, Toronto. Standardizing language and reporting data so that it's not only accessible but comprehensible to diverse groups within a company can help reduce misunderstandings, eliminate duplicative research efforts, and increase innovative ideas by constantly bringing in fresh perspectives.
» Jump to indexing (document details)
Full Text
(2004 words)
(c) 2008 Dow Jones & Company, Inc. Reproduced with permission of copyright owner. Further reproduction or distribution is prohibited without permission.
In the quest for innovation, companies continuously collect and analyze ever-increasing amounts of customer data. But what if a company has already collected the most valuable information and doesn't realize it?
Sometimes the answer to the innovation challenge lies not in collecting new customer data but rather in finding creative ways to share and repurpose customer information that's already in-house.
Consider two quick examples. First, gambling giant Harrah's Entertainment Inc. has long collected customer information to optimize marketing offers and increase customer loyalty. But eventually, Harrah's was able to use those details about the spending habits of its high rollers to help it redesign its casinos, positioning games to encourage those customers to play even more.
Second, Royal Bank of Canada uncovered a new innovation engine, and gave its consumer-credit business a boost, by looking again at old rejected loan applications. The idea was to see whether any of those applicants had improved their credit ratings in the meantime, and then offer credit cards to those who had.
None of this is as obvious -- or as easy -- as it may seem. The transition to greater sharing of and finding new uses for existing information can be difficult when each department is used to collecting and analyzing its own data, often with a kind of tunnel vision that excludes the possibility of unexpected uses and discoveries. At Royal Bank of Canada, the decision to look again at rejected loan applications was a dramatic switch for the risk- management department, says Cathy S. Burrows, director of client strategy at RBC Insurance, Toronto. Says Ms. Burrows: "To have a risk group saying, 'Let's go out and go after customers we have turned down and offer them money' -- that is a huge change."
We studied companies that successfully share and repurpose data in ways that lead to improved operations, new products and greater revenue. Information cross-leverage can produce a fountain of new ideas across disparate domains within a firm. Indeed, our research revealed there is a strong correlation between a company's ability to share and find new uses for information and its capacity for innovation. Firms with these skills are also more likely to complete information-intensive projects less expensively, more quickly and with greater success.
Some of the companies we studied accomplished information cross- leverage systematically, with companywide routines and lots of encouragement from the head office. Others lacked set procedures but still achieved such innovations thanks to fortuitous circumstances -- and having the right organizational outlook when it comes to managing data.
What both groups had in common was a handful of practices and values essential for any company that wants to tap existing data as the basis for a new stream of innovation. What follows, based on our research, are three principles that lie at the foundations of successful information cross-leverage.
The Power of Standardization:
Everyone Must Speak the Same Language
A lot of valuable information is wasted in companies because employees don't describe or understand information in the same way across the organization. Different measures may be used to tackle the same assignment, leading to invalid comparisons. Even such words as "profit," "overhead" and "attrition" can mean different things in different departments, units and regions. There is also a tendency for groups to customize data gathering and reporting methods for their own needs and interests, without regard for whether their results would be useful elsewhere in the company.
Standardizing language and reporting data so that it's not only accessible but comprehensible to diverse groups within a company can help reduce misunderstandings, eliminate duplicative research efforts, and increase innovative ideas by constantly bringing in fresh perspectives.
Best Buy Co., Harrah's and Royal Bank of Canada all make a systematic effort to share information across departments and functions and to standardize terms. At Best Buy, standardizing certain customer information led to a host of marketing, merchandising and operational improvements. Transaction data originally collected to learn how much Best Buy customers typically spend went on to help determine what products to stock, and at what prices. The data were even shared with suppliers to build higher-value products and services customized for Best Buy.
Creating -- and enforcing -- standardized language and measures requires the significant involvement of top management. Robert Willett, chief executive of Best Buy International and enterprise chief information officer of Best Buy, says the effort of standardization at Best Buy was "ginormous and not for the faint- hearted." When he joined the company nearly four years ago, it had "400 or 500 different ways of measuring things. . . . I said, 'I will not develop another piece of technology until there is common agreement and we do not have every part of the organization wanting a different measure.'"
The company had to arrive at a single interpretation on a host of terms, measures and definitions used across the organization, Mr. Willet says. "I brought together folks from all parts of the organization, drove it like mad for 10 months, and then we began to realize value from it," Mr. Willet says.
Royal Bank of Canada, for its part, invested heavily in developing an online, wiki-style dictionary that allows its workers not only to find definitions of terms used in their business, but to modify those definitions if they are unclear, and to provide additional knowledge about the terms and contexts for their use.
"It is always a tough economic decision in the economic-planning process to point out what the value of information standardization is," says Amos Khim, managing director of corporate forecasting at UAL Corp.'s United Airlines. "But if it is done, it is a big benefit."
For example, at United, he says, standardized reporting of ticket sales has helped operations managers better plan staffing needs at airports not just according to seasons, but down to the busiest times of day. "If there are multiple sources of ticket booking information, then I will get conflicting information," Mr. Khim says. "Two people writing different reports may have a totally different view or outlook of what is happening . . . because they are looking at different [nonstandardized] information sources."
The Value of Abundance:
Give More Information Than
Was Sought
While it's important for colleagues to be able to share what they know, it can be just as important to explain how they know.
People seeking information need not only answers to their questions, but knowledge that puts the answer in context and promotes a broader and deeper understanding of the subject. They should know how the information was collected, what its strengths and weaknesses are, and what its intrinsic value is.
Much of this doesn't happen in most organizations without a determined effort. Royal Bank of Canada, for one, makes a point of not only sharing information across departments but also providing clear and detailed documentation of data, explaining its relevance and providing supporting detail. To see how helpful this can be, consider the following:
Several years ago the bank created an automated system to determine which customers should receive a courtesy overdraft when they had just written a check for which they had insufficient funds. The system drew on the customers' history of overdrafts and repayments, their credit scores, an analysis of their value to the bank and risk factors. Afterward, an ATM product manager read the notes about the system, which included details about how it made decisions, and proposed something similar for ATM customers. Now, for a fee, Royal Bank of Canada offers credit to approved customers who wish to make an ATM withdrawal that exceeds their account balance. James Moore, director of client decision management, says the project "cost less than $100,000 to build and has earned millions for the bank."
A lot more sparks can occur "when you have that many more analysts who can access the information and apply it to the problems that they are experiencing," says RBC Insurance's Ms. Burrows. It allows potential reusers to spend most of their time looking for insights and new ways to leverage existing information, rather than gathering new data, trying to understand it and molding it into something useful.
The Virtues of Generality:
Don't Make Questions Too Specific
Sometimes it's not enough to change the way information is presented. Employees may have to change the way they ask for information, too.
Queries seeking data should pose general questions that reveal the broad goals of the undertaking. This helps other departments recognize when they have data relevant to the task. People who initially gathered data for a different purpose often aren't aware of how that information might help with a new project.
At Harrah's, for example, managers trying to figure out which hotels were in need of expansion posed their underlying question as "What is the demand for our hotel rooms?" The managers might have circulated a request for occupancy rates at the various hotels instead. That would have been simpler. But posing the more general question netted both occupancy rates and numbers of people unable to book a room, either because the hotel was full or because the customer was unwilling to pay the current rate. In the end, posing the question in a general fashion gave the managers an even richer field of data on which to base conclusions.
As Best Buy's Mr. Willett observes: "It is more important to understand the context of a problem and not just the answer. If you help me understand what you are trying to achieve, if I understand what you are trying to get to, then we are pretty smart about figuring out how to get there."
Broad questions speed up learning curves by exposing the researchers to more kinds of information. They also can turn up answers to problems that the researchers weren't even thinking about. Employees in the London office of Continental Airlines Inc., for example, asked the company's corporate IT staff to provide them with customer data that could be used to predict staffing needs based on ticket purchases. This broad request ultimately netted data that also helped the London group automate a task formerly performed by hand: determining which Continental passengers were required to pay a U.K. departure tax slapped on travelers last year, a tax charged to passengers who stay in Britain longer than 24 hours.
Getting employees to ask more general questions isn't easy. It requires a conscious abandonment of short-term efficiency in service of greater effectiveness. Indeed, many organizations take individuals to task for being "inefficient" when they ask general questions, which are exploratory and require more resources, even though the long-term benefits often outweigh the costs.
Some companies hold brainstorming sessions to get employees to make broader inquiries and increase their use of existing data. Attendees may be asked such simple, open-ended questions as, "What else could we do with this data?" Or, in even broader terms, "What is likely to be out there?" instead of, "What do I need?"
---
Dr. Gulati is a professor of business administration at Harvard Business School, Harvard University, Boston, Mass. Dr. Oldroyd is a professor at SKK Graduate School of Business in Seoul, Korea. Dr. Puranam is a professor at London Business School and co-director of the Aditya Burla India Center at the school. They can be reached at [email protected].
---
For Further Reading
These related articles from MIT Sloan Management Review can be accessed online at sloanreview.mit.edu/wsj
How Do They Know Their Customers So Well?
By Thomas H. Davenport, Jeanne G. Harris and Ajay K. Kohli
(Winter 2001)
Many companies know about their customers, but few know the customers themselves or how to get new ones.
How to Profit From a Better 'Virtual Customer Environment'
By Satish Nambisan and Priya Nambisan
(Spring '08)
Many companies have established technology-based platforms or virtual customer environments that allow them to partner with their customers in innovation and value creation.
Delight or Despair
By M. Eric Johnson
(Summer 2007)
The ability to harness customer data is proving to be a significant source of competitive advantage for service industries.