|
|
|
|
The Business Value of Data by John
Hagel Does
Data Have Value? Does Data Have Value?Data is the most undervalued asset of business. Endless treatises have been written about the business value of information and knowledge, but one has to search far and wide for any discussion of the business value of data. When the topic comes up, more likely than not, it is dealt with in dismissive tones as in a recent report from a well-known analyst firm (name withheld by the Editor), that declared: “the recommendation to clients is to recognize that data is worthless.” In a sense, of course, this is correct. Data without context and without understanding of how it can be applied to business has little value. Even in this form, however, data has option value. It is a raw material that, with appropriate context and understanding, has the potential to generate significant business value. One can endlessly debate the distinctions between data, information and knowledge. For our purpose, let us consider data to be the “facts” that describe our business. For example, we produced 7,000 widgets yesterday. Information puts this data into context and creates meaning for the receiver of the data. For example, our daily capacity is 25,000 widgets and on a normal day we produce at least 20,000 widgets, so this production level is a significant shortfall relative to our daily performance. Knowledge is much more complex. It is tied to individuals who evaluate and assimilate information, based on their own frameworks of experience and understanding, and who then make choices about how to use the information. For example, the plant manager may know that the problem is with a new assembly machine that could not handle a certain component and led to a spike in reject rates for finished goods and that the only way to quickly deal with this is to sideline the new assembly machine until the troublesome component can be replaced (it turns out the data only reports finished widgets that meet quality control standards). Data is like any other raw material, like iron, silicon or crude oil. Very little can be done with these materials in their raw form. They all need to be processed to generate business value. Yet, even in this raw form, they all represent significant assets. The revenue that can be generated from finished goods using these materials determines their value. In fact, the quality of the raw material matters a lot – if you have low-grade crude oil and you will find that it has only limited business value because it will generate a very modest amount of gasoline and other high-end petroleum products. Assessing the value of data depends on a clear understanding of the information, knowledge and action that it can generate. The value of the data will also vary significantly depending on its quality – fragmentary, inaccurate, irrelevant or less timely data will generate less business value relative to higher quality data. One of the reasons that data has such a bad reputation is that the link between the data and its potential business value is poorly understood. This is why we are so often confronted with either data overload or data draught. In the absence of a clear link, business will tend to either overproduce or under produce data. Rapid improvements in the price-performance of information technology create a bias towards overproduction of data. The paradox is that we are drowning in terabyte seas of low quality data while we are desperate for high quality data to drive high value business decisions and actions. It is perhaps no wonder that data has such a poor reputation these days. Under these conditions, data represents a liability rather than an asset – we invest millions (in many cases, billions) of dollars to produce and store data and, more often than not, generate a negative return on investment. Potential Value of Data IncreasingHere’s another paradox – at the same time that businesses are destroying economic value through overproduction of low quality data, market forces are dramatically increasing the potential value of high quality data. We used to live in a business environment where structural advantages – scale, location, regulatory barriers, etc. – determined competitive advantage and financial performance. In this environment, data played a relatively modest role in generating business value – it was essential for record-keeping and reporting performance, but it did not play a significant role in creating structural advantage. This has all changed. Over a period of decades on a global scale, two long-term trends have converged. IT innovation and public policy shifts towards open markets have rapidly eroded traditional structural advantages. In their place, superior insight and execution have become the keys to economic value creation. Data plays a critical role in enhancing the potential for superior insight and execution. As markets and industries become more dynamic, access to high quality data becomes essential to anticipate potential changes on the business landscape. Similarly, high quality data plays a key role in rapidly improving operational performance. Without tight feedback loops on operational performance, managers have a much harder time identifying gaps in performance and determining whether specific changes to practices or processes produce desired improvements in performance. One of the most powerful ways to improve operational performance in dynamic markets is to compress operational cycle-times. One of the key bottlenecks in these efforts is lack of access to timely and accurate data. Similarly, as customers become more powerful in dynamic markets, they require access to timely and accurate data from the vendors they deal with. Conversely, vendors can gain significant business advantage through privileged access to data about customer needs. As financial markets become more competitive, companies confront increasing pressure to report results using high quality data. Investors will tend to favor companies that can stand by their numbers and avoid unpleasant surprises. Predictability in reporting in many cases will have higher value to investors than superior operational performance inadequately reported. High quality data can generate significant rewards in more competitive and dynamic markets. But that is only half the story. Low quality data generates increasingly severe penalties. Companies that fall behind in terms of insight or execution will find themselves experiencing severe margin squeeze and lower growth rates – and financial markets will punish them harshly. Financial markets also punish companies that “surprise” investors with unexpected shortfalls in performance or restatements of financial performance. That punishment is bad enough, but the punishment may be even more literal, in the sense of civil or even criminal penalties for failure to meet increasingly strict reporting requirements to a broad array of regulatory agencies around the world. Now, to clarify once again, it is information and knowledge that ultimately generate these rewards or penalties for business, not data in its raw form. But, if the data is not available or is low quality, the information and knowledge will be more limited and the business value will be diminished accordingly. Data in its raw form cannot generate business value, but it is the foundation for generating the information and knowledge that does create value. We need to pay much more attention to data as the pre-requisite for business value creation. As we have seen, the potential for value creation through better quality data is rapidly increasing. The sad fact is that, for most companies, the gap between potential and reality is growing wider over time. The challenge for all executives is to work to close this gap. Those who do will reap the rewards. Those who don’t will suffer the penalties. All Data is Not Created EqualTo close the gap between potential and reality, we need to understand the different categories of data and the shifts in value among these categories of data. Understanding
categories of data
These five data types can also be categorized in terms of whether they provide financial or operational views of the business environment. For example, data regarding business processes might report on operational elements like the number of units processed on a particular assembly line or on financial elements like the cost to run the assembly line each day.
Significant current investment seeks to provide more unified data views of activity at the enterprise level. We are all familiar with major investment initiatives to provide integrated views of customer activity that bridge beyond individual products or specific sales channels. Many companies are also faced with the challenge of integrating operational data across manufacturing plants, distribution centers or geographic business units. But the scope of data views need not stop there. Most core operating processes extend across other enterprises — not just direct business partners, but often across multiple layers of business partners throughout an entire value chain. Customer and market data by definition extend beyond the enterprise and here the scope of data views can expand as well. In terms of customer data, it may be necessary to understand the “share of wallet” or even “share of stomach” that a specific company represents, requiring a view of the total purchasing activity of individual customers in specific product categories, rather than just focusing on the total purchases of the customer from an individual company. The scope of broader market data may grow as well as companies begin to realize that traditional industry boundaries are eroding and that potential competitors and business partners may come from very unexpected sources.
To create additional value, companies need to innovate continuously in terms of both products and processes. Some of this innovation will be focused on reducing product and operating costs, but the real opportunities for value creation come from discovering new ways to deliver additional value by addressing a broader range of customer needs. Once again, data is an essential input into the innovation process, either to support cost reduction or customer value enhancement. To support the innovation process, companies often focus on leading indicators in the business environment. These leading indicators can help management to anticipate emerging opportunities and challenges. For example, trends in the curriculum and enrollment rates in engineering schools can provide early indication of potential skill gaps or surpluses that will shape employment patterns in the high technology industry. Effective data gathering constitutes an essential foundation for monitoring these leading indicators. Data also plays a critical role in supporting the external reporting requirements of a company. To be successful, companies must communicate effectively and reliably with investors regarding the financial performance of the business. Ultimately, operational performance will shape financial performance, but high quality data to support investor communications can have a significant impact on the margin in terms of building investor trust and confidence. Companies also have a broad range of reporting obligations to a variety of fiscal and regulatory agencies of the government and the quality of data will determine how effectively companies meet their reporting obligations. Shifting Value of Data CategoriesIn our dynamic business world, it should not be surprising that we are seeing significant shifts in terms of the value of different data categories. Let’s highlight three broad shifts that are re-shaping data strategies of leading edge companies. Expansion of data focus. First, there has been a broad expansion in the focus of data from internal to the enterprise to external domains. Several forces are driving this expansion. First, customers are increasing in power relative to vendors, so it is becoming more critical to understand customer needs and behavior. Second, businesses are becoming more dependent on each other to deliver value to customers. Whether this takes the form of outsourcing, strategic partnerships or lean supply chains, it is becoming increasingly important to gain access to high quality data from a wide range of business partners. Third, as competition intensifies in markets around the world, companies need to obtain data regarding both actual and potential competitors. Of course, this expanding focus does not mean that internal data is diminishing in importance. The real challenge for management is to effectively integrate internal and external data to provide a broader view of business activity. If the effort to integrate data across departments and functions within enterprises has been challenging, it is a modest effort relative to the growing complexity of integrating data from multiple sources across enterprises and customers. Increasing value of certain types of operational data. In more stable industrial markets dominated by structural sources of advantage, the primary emphasis was on financial data. Keeping score was the primary concern of managers. As these structural advantages have eroded, management attention has become much more concentrated on operational data. Financial data is ultimately a lagging indicator – it tells you how you did in the past, but offers little insight into what management decisions need to be made to improve financial performance in the future. Margins may be eroding, but without solid operational data it is hard to determine whether management needs to focus on more effective marketing, increasing sales skills, improving manufacturing plant productivity or redesigning the product for greater manufacturability. In general, operational data can provide important internal leading indicators of improved financial performance. As suggested above, the expanding focus of data leads to the increasing importance of external types of data like customer and market data. But “internal” types of data like product or service, process or practice and resource data all become much broader in scope, extending beyond the boundaries of individual enterprises and requiring views of operational data within multiple enterprises and customers. Growing importance of data to support innovation and anticipate market events. As competitive pressures intensify and market uncertainty increases, the purpose of data collection is shifting as well. The early response to intensifying competition is usually to increase coordination and problem solving efficiency. This helps to reduce cost and increase competitive strength. But cost reduction, especially in this form, while essential, offers limited value. As competition increases, cost savings tend to get competed away and captured by customers. The only way to create significant and sustaining value in competitive markets is to focus on continuous innovation to deliver more value to customers with fewer resources. As a catalyst for business innovation, data regarding unmet customer needs, complementary resources from other companies and innovative practices in other industries become particularly valuable. In highly uncertain markets, those who have privileged insight regarding future market trends possess a significant advantage. Hence, high quality data regarding leading market indicators becomes particularly valuable. Often, these leading indicators are in the “periphery” of the business (e.g., marginal customer segments or adjacent industries) precisely because they are not yet sufficiently important to today’s business performance. As a result, even knowing what data to seek or how to obtain it can be very challenging. Technology Enablers Enhance the Value of DataTwo factors determine the value of data: the business value that can be generated from the use of data and the cost of making data accessible in useful forms to appropriate business users. The latter cost covers many different activities – generating or acquiring data, storing it, distributing it, aggregating it and analyzing it. Technology is playing a significant role in driving down the cost of all of these activities. As the cost of making data accessible declines, the overall value of specific data will increase. Technology is key to reshaping the costs of data use. We are all familiar with Moore’s Law that has accurately predicted the relentless doubling of computer processing power every 18 months. This is systematically reducing the cost of data processing. Similar, but even more powerful, trends have been reducing the cost of data storage and data transmission. The price performance of data storage technology has been doubling roughly every 12 months and the price performance of data communication technology has been doubling at an even faster rate – roughly every nine months. Other technology trends are also having a profound impact on lowering the cost of data use. Web services technology is beginning to play a critical role in reducing the cost of data aggregation. By providing standardized formats for representing data, Web services technology will help businesses to gather data from many different sources and to automatically aggregate it in a comparable form. Web services also play a key role in facilitating data transmission from one area of the business to another. Today, businesses consume significant expense in moving data across technology platforms. In many cases, the data must be printed out from one system and manually re-entered into a second system. Web services technology helps to automate these connections, and is proving especially valuable in supporting automated data transfer across multiple enterprises. More broadly, as digital technology expands its reach, displacing both analog machinery and human muscle, it generates more and more data as a by-product of other activity. As one small example, the automobile is becoming a significant archive of data generated by the growing number of digital circuits controlling various dimensions of the automobile’s operation. The lower cost of digital technology is also making new forms of data capture economically viable. RFID technology illustrates the potential to make “dumb” objects like packages “smart” and, in the process, generate a flood of new data to help businesses create more value. The growing amount of data available to business creates its own challenges. How can business quickly process this data and generate information that is useful and timely to its employees? The awareness of this challenge may help to explain the increasing interest in business intelligence and business analytics software. Although these two categories are often used interchangeably, business intelligence software generally focuses on generating standardized, pre-configured reports focused on providing people with information about the historical performance of the business. Business analytic software, on the other hand, helps analysts to discover new insights on the business by exploring potential relationships among data sets. For example, by comparing usage patterns among cellular telephone users and subscriber termination patterns, it might be possible to identify early indicators of subscription termination. Certainly any technology that helps business users to extract more information and insight out of the growing quantities of data will be a welcome addition in the continuing effort to generate more value. Technologists often paint grand visions of an automated world where data flows smoothly but people are absent. Some of the value of data can certainly be realized through more automated processes. But a substantial portion of the value will remain untapped unless technologists focus on the ways data can be more effectively used to amplify the efforts of the people in the organization. The gap between the potential of the technology and the actual value realized from the technology is growing. We will not close the gap until we concentrate on building the skills required to use the data and the technology more effectively. Developing an Action Plan to Increase the Return on DataData is like any other business resource — it can be a source of value creation or value destruction depending on how it is managed. To create value, management must systematically focus on the economics of data. How can we maximize the economic value of data? How can we reduce the costs of data use? To develop a high impact data strategy, it is helpful to use a FAST approach, simultaneously managing four streams of effort — focus, accelerate, strengthen and tie together. Impact can be easily lost if executives fail to effectively balance long-term data strategy with the need for near-term business impact. If the balance is lost, the company will either drown in a sea of data or weaken because necessary data is not available. The FAST approach helps to maintain this delicate balance. Focusing
on high impact data How to do this? Some relatively simple techniques can help. First, work with the senior management team to clearly define a few “closely watched numbers” that have a particularly significant impact on business performance. These numbers may focus on key economic leverage points of the business. For example, in consulting operations the utilization rates of the consulting staff are particularly significant in determining financial performance. For companies embarked on a major new strategic direction, closely watched numbers may focus on key operational milestones in major new business initiatives. For example, a company making a transition from analog to digital products like Kodak might focus on the ratio of digital cameras sold to analog cameras sold. Few senior management teams have explicitly agreed upon these closely watched numbers (if there are more than ten for the entire company, then they are unlikely to be closely watched) and communicated these to the organization. This discipline will not only help to build an effective data strategy. It will also help to focus the entire organization on the performance that really counts. Before locking into these closely watched numbers, it may be helpful to ask the senior management team to answer three basic, but very difficult, questions. What will our markets look like 5-10 years from now? Given this, what kind of business will we need to have to continue to create economic value in these markets? What measures can we develop to monitor our progress towards this new kind of business? Without this explicit effort, the closely watched numbers may end up focusing too heavily on today’s business and distract attention from the changes required to remain competitive. Another technique is to ask senior management to identify the pivotal jobs within the company. Pivotal jobs refer to the roles that have a disproportionate impact on the economic and competitive performance of the company. These pivotal jobs will obviously vary by industry and company. In insurance, risk assessment is likely to be a pivotal job. In systems integration companies, project managers are likely to hold pivotal jobs. Once these few pivotal jobs have been identified, work with seasoned occupants of these jobs to understand what their key unmet data needs are and ask them to prioritize these unmet needs in terms of impact on their job performance. 80/20 analyses can also provide a useful focusing tool. Work with senior management to break apart the business on multiple dimensions, asking the 80/20 question: which 20% of the business generates 80% of the profits? Apply this question to the company’s, business units, product portfolio, customer segments, distribution channels, geographic units and operating processes. You’ll be surprised at how broadly the 80/20 patterns apply. It will also dramatically simplify the task of identifying the data that is most important to the business. Once you have identified the most profitable parts of the business, work with the senior executives in these areas to assess unmet data needs. These techniques so far are designed to focus attention on the data with the highest potential for value creation. In addition, it is important to work with senior management – especially the CFO and chief legal counsel — to identify the greatest data risks to the company. Where could the absence or misuse of data cause the greatest harm or destroy the most value for the company? Often these risks will be in the financial reporting and regulatory domains, but there may be issues as well regarding competitive vulnerabilities or operating risk assessment. All of these techniques help to identify and prioritize the highest impact data required by the company, but they do it in a very focused way that avoids the risk of becoming consumed in assessing all possible data that might be relevant to the company. Accelerating
business impact One or two significant near-term initiatives should be launched in parallel to reduce the cost of existing data capture and storage programs. Just as most companies are missing key data required to improve business performance, most companies also are gathering data that is not being used or that is being stored inefficiently. Gartner reports only 30% of the utilized data in the average corporation is actually active, in terms of being accessed within the last 90 days. At the same time, storage costs are increasing by around 30% per year, but only 62% of enterprise storage is effectively utilized. Clearly, there are ample opportunities in the average corporation for substantial cost savings through more efficient data management processes. Like the initiatives above, each of these significant cost saving initiatives should be supported by an explicit business case before it is launched. Strengthening
capability Tying
together It is critical to assess the actual impact of data improvement initiatives relative to their expected impact. According to the market research firm Giga, only 5% of all IT projects are subjected to a retrospective to learn from the experience. Any unexpected discrepancies should be evaluated relative to the focusing analyses that led to the selection of the initiatives in the first place – can the analyses be refined to ensure that these discrepancies do not reoccur? Similarly, unexpected discrepancies may suggest refinements to capability building initiatives to accelerate and increase the realized impact. As the efforts to identify and prioritize high impact unmet data needs proceed, existing data improvement initiatives might be reconfigured and new data improvement initiatives might be developed. These refinements to the data improvement initiatives may in turn suggest higher impact capability building initiatives. The key is to continually assess progress across all three streams and to be alert to opportunities to refine efforts in other streams based on new learning in any individual stream. The FAST approach is ultimately a framework for accelerated learning and performance improvement. It will help companies to realize more value from their data more quickly. The IT department cannot execute this approach alone, but it provides a vehicle for building close coordination between the IT group and the rest of the corporation. The Bottom LineAs corporations seek to compete more effectively in rapidly changing markets, data is a critical raw material that will increasingly shape corporate performance. Capturing the value of this raw material will require a detailed understanding of its sources and the technologies available to process it and deliver it to the appropriate people. As this resource becomes more plentiful and lower cost, the real source of advantage will accrue to those companies that understand its true value and act aggressively to harness this value. The FAST approach provides a useful methodology to build shared understanding of the value of this resource across the corporation and to focus efforts on the highest impact use of data. The winners in the competitive battles ahead will be those who succeed in maximizing the return on data. Copyright © John Hagel 2003. All Rights Reserved. -- John Hagel is a business consultant and author. He works with senior management of large enterprises around the world to shape business strategies and improve business performance. His experience includes senior management positions in technology businesses and sixteen years as a consultant with McKinsey & Co. He most recently is the author of Out of the Box, a book on the business implications of Web services technology. He is also the author of the best-selling business books, Net Gain and Net Worth. In 2002, Accenture identified John as one of the top 100 business "gurus" in the world. In 1999, Business Week named John one of the e.biz 25 most influential people in electronic business. John has also received two McKinsey Awards for best articles published in Harvard Business Review. He has been designated as a Forum Fellow by the World Economic Forum. John received his M.B.A. from Harvard Business School and a J.D. from Harvard Law School. He also has a graduate degree (B.Phil.) from Oxford University, as well as a B.A. from Wesleyan University. More information is available at www.johnhagel.com. |
sponsored by
|
||
| Copyright (c) 2004-2005, nextslm.org. All Rights Reserved. Legal Statement. | |||
