Balanced Scorecard for R&D
Not long ago the average pharma development organization spent $100 to $200 million a year. It consisted of about 100 people working on only a half dozen compounds. Its leaders typically knew all their employees by name. There was a good deal of visibility within the organization: Managers could see from top to bottom and could make decisions aided by nothing but their own brains.
In the past ten or 15 years, much has changed. Each of the top ten pharma companies spends more than a billion dollars a year on development. A development organization today consists of 1,000 or more people and works on dozens if not hundreds of compounds. The visibility of a decade ago has vanished. Basic decisions-setting priorities, allocating funds, deploying and re-deploying resources and so on-have become extraordinarily difficult.
If there is a top-down visibility problem, there is also a corresponding bottom-up problem. From the operational level, it is often difficult to see strategic vision and business objectives, or at least to see how they connect to a department's day-to-day activities. To senior management, corporate strategy may be vivid and urgent, but too often the people who need to execute the strategy see it as something abstract and irrelevant.
The results for R&D are predictable: a crisis in productivity that has left most companies scrambling for new products. Many factors contribute to the crisis, but one key to solving it is better management. Companies need a system of accountability so that each component of R&D has disciplined, focused operational objectives that support corporate strategy and enable decision makers to identify and support rewarding avenues. However, that is not as simple as it sounds. Mid-size and large pharmaceutical companies typically address multiple therapeutic areas, have globally dispersed R&D facilities, and must work with large networks of clinical coordinators, physicians, and patient groups around the world.
What is more, until recently, R&D managers have found it almost impossible to structure and analyze the enormous volumes of data their departments generate. Paper reports, the management tool of choice for most companies, are too slow and do not permit managers to work directly with business data across the operation. Moreover, these reports address only results, giving management little or no chance for early intervention.
During the past few years, however, powerful tools have become available to structure and analyze large volumes of data for managers. One of the most promising developments is the emergence of "balanced scorecard" software. Made by several dozen suppliers, balanced scorecards are both a means of processing data and a tool for implementing a management philosophy. Though they have only recently been applied to pharmaceutical R&D operations, balanced scorecards are well suited to the industry's needs.
A balanced scorecard system is designed to achieve several goals:
This article describes the balanced scorecard system implemented in the medical development operation of Pharmacia. The system has provided greater transparency for managers, and though intended for long-term results, it has had impressive short-term effects. In the past two years, Medical Development introduced several process redesign initiatives that could be visualized in a dynamic way using the balanced scorecard approach. Cycle times on critical path in clinical trials were shortened dramatically, by 34-75 percent; per patient costs were reduced by 5 percent; and the number of patients recruited per clinical research site manager increased by 34 percent, while maintaining quality. Although it did not bring about these astonishing improvements directly, the scorecard supports and contributes to the overall implementation and ongoing decision making.
A Structure of Indicators
The balanced scorecard concept grew out of research done at KPMG in the early 1990s. Beginning in 1992 it was introduced in an influential series of articles in the Harvard Business Review by Harvard business professor Robert S. Kaplan and consultant David P. Norton. The two authors explained that traditional accounting was not up to the task of analyzing the health of complex contemporary businesses. "Ideally," they wrote in their book The Balanced Scorecard, "this financial accounting model should have been expanded to incorporate the valuation of a company's intangible and intellectual assets, such as high-quality products and services, motivated and skilled employees, responsive and predictable internal processes, and satisfied and loyal customers."
Most important, traditional accounting focused on lagging indicators: outcome measures such as sales figures that showed what a company had done in the past but not what it would do in the future. Kaplan and Norton wanted to mix outcome measures with "performance drivers": leading indicators such as manufacturing cycle times that suggested where the business was headed. And these indicators needed to be linked to overall business strategy.
To accomplish that, Kaplan and Norton argued that managers should track a company's performance in four main areas:
They encouraged managers to build a logical structure that linked measurable indicators to corporate goals in each of those areas. That structure, which often entailed thousands of performance indicators organized under several dozen goals, was the essence of the scorecard.
It took a few years for the balanced scorecard concept to become a success. In part, that was because, for the approach to work, companies need the ability to assemble vast amounts of data in data warehouses and deliver the information to managers' desktops. Today, the term "balanced scorecard" is as likely to refer to a software product as to a logical structure, but both are necessary.
On one hand, a balanced scorecard is simply part of a software chain that moves data from corporate databases onto the executive's desktop. A development organization typically has many databases. At Pharmacia, for example, the list included, among others:
Other sources of data can be easily added to the system, according to business needs.
On top of the source databases sits a data warehouse, which collects and indexes information. The ideal is one warehouse for all of a company's data. In practice, it is often necessary to use several warehouses. Typically, only experts and analysts possess the skills to work directly with warehoused data. Other users need a software tool that sits on top of the data warehouse to extract, structure, and organize the information.
What makes balanced scorecard software unique, though,is not the fact that it moves data from the warehouse to the desktop but how it organizes the data. Instead of confronting raw data, executives can see information that is already plugged into a structure of corporate goals and benchmarks. And that is perhaps easiest to understand by experiencing the scorecard the way a manager would experience it.
When Pharmacia's balanced scorecard program (created by Oracle) is launched, it displays an executive summary screen. (See "The Big Picture," page 86.) On the screen are five basic headings, slightly modified from Kaplan and Norris's original four:
Under each heading are either four or five operational objectives, for a total of 23 (a number that Norris and Kaplan called optimal). Under the heading Process Excellence, for example, there are five objectives:
Each objective is flagged with a traffic signal icon: green if the R&D department is meeting or exceeding its targets, yellow for borderline performance, and red for objectives the department is failing to reach.
The opening page is both a dashboard that gives current information on performance and an index to thousands of specific numbers that underlie it. For example, if a manager sees a red icon marking "optimize trial performance-timeline" and clicks on it, he or she will be offered a menu of the key performance indicators (KPIs) that it represents.
Each KPI is aligned with the business objective. Click on a KPI, and it is possible to view the numbers in several different graphic representations. (See "Data in Depth.") Pharmacia's balanced scorecard system calculates about 20 KPIs, and most can be broken into several measures. For example, clinical trial cycle time is a KPI, but a manager who wants more detail can click through to its individual components: study set-up cycle time, patient recruitment cycle time, report writing time, and so forth. Pharmacia's balanced scorecard tracks about 2,000 measures in all.
The goal in designing such a system is to build up a logical relationship among KPIs. The ideal would be to set up mathematical relationships, but that is not often possible. Even the relationship between clinical trial cycle times and the overall project cycle time in bringing new products to market is only semi-mathematical.
Other goals are much more difficult to measure. For example, an R&D organization needs to train its people and develop its leadership team. It is possible to set up some measures that track progress toward those goals: One might decide that the drug development effort would benefit if everyone in the department were fluent in using Microsoft Excel and then set up milestones for getting 15, 50, and 100 percent of personnel trained. But the connection to the corporate goal of bringing drugs to market is much less direct.
Attrition, Cycle Time, Cost
Pharmacia's balanced scorecard is designed to focus managers' attention on three aspects of the R&D process:
Attrition rate. Of the large number of compounds that enter the development process, only a small fraction make it through. Attrition rate as a measurement shows how many compounds are needed at any point in the development process to achieve a target number of successful registrations. The scorecard provides detailed attrition rates for every phase of the development process.
Cycle times. Project cycle time is the amount of time it takes for one compound to make it through the entire development process from first testing in humans to approval. The balanced scorecard also provides a detailed breakdown of cycle times for individual phases of development.
Financial funding. Funding is more or less fixed on an annual basis, but it is an important limitation on how many projects a development organization can afford to finance.
With those three factors, a manager can calculate how many compounds are needed at each development stage and identify gaps. That is a dynamic process. On a quarterly basis, cycle times change and attrition changes, so the gaps also vary. But occasionally there will be a huge, obvious gap, and then the availability of complete, current data makes it easier to come up with an appropriate strategy. If there is a shortage of compounds in an earlier stage, for example, there is the option of licensing a compound from outside or of accelerating targets in the discovery stage. Either of those solutions would cost tens of millions of dollars but would still be a substantial saving compared to the hundreds of millions or billions of dollars it costs to license a compound in Phase III.
Better Than Paper
Compared with the more widely used metrics report on paper, a balanced scorecard offers several advantages:
In today's R&D units, senior managers get paper reports from professional analysts. The analysts often end up scrambling to verify data extracted from different databases. Operational people, though spending time and effort entering information in databases, don't get much back in return.
Through a balanced scorecard system, people will get feedback from what they provided almost in real time. It is in their own interest to maintain the quality of the data and keep them up to date. Moreover, they have the opportunity to communicate with their colleagues, senior management, and the balanced scorecard manager through electronic forums that are part of the system.
The balanced scorecard approach enabled Pharmacia's medical development organization to set up a series of internal and external performance metrics to measure adherence to strategic objectives and the operational requirements to support them.
By using this dynamic measurement tool and receiving an immediate indication of performance, senior managers can be more results-driven. At any time, they can see what's happening in the pipeline, the status of each project, its staff and resource costs, and whether it is meeting its milestones and its forecast deadlines. The benefits of any process reengineering can be seen visibly and quickly, thus ensuring that the organization realizes value on its investment. Furthermore, a feedback loop from the scorecard outputs enables the senior management team to adjust targets and priorities in an ongoing, focused manner. In this way, the management team can start to use data from many sources to influence its strategies, targets and decision making and thereby improve its overall business productivity.
A word of warning, though: There may be initial resistance to the idea of bringing greater accountability to R&D. Line functional heads are accustomed to running their own show, and may not be comfortable with such transparency of performance. But if the scorecard is used in a positive, constructive manner to support business decisions, give greater focus to key goals, and improve uptake of new processes, managers will come to appreciate its benefits.
How Digital Medicine Can Pinpoint Dosing Regimens to Optimize Drug Efficacy and Safety