Chapter 3 - Action, Personnel, and Cultural Controls In early 2016, Frank DeMartino, senior vice president of Science and Technology at \" Pharmaceuticals, Inc., reflected on his concerns about the challenges his company faced in controlling its research and develop- ment (Ram) activities: Fi&D is the most critical part of EyeOn's business. The company will thrive only if we are effective at developing new breakthrough products. In manag- ing the research function, we have to address three difficult but important issues. First, we have to decide how much to spend on non. Then we have to decide how to allocate the resources among the various programs and projects. And, nally, we have to ensure that the resources are used effectively. How we address these issues determines how productive our research activity will be. I am especially concerned about the third issue how to control the use of our resources. I don't think we do a very good job of measuring our pro- ductivity. At the time we are spending our resources, both money and time, and even for some time after they have been spent, it is very dif- cult to tell how productive we are being and have been. We could be missing some important infor- mation about problems we might be having. I feel we should do some thinking about this issue and what we can do to improve the tracking of our Fl&D productivity. The company and its products EyeOn Pharmaceuticals, Inc. (EyeOn) focused on the ophthalmology segment of the pharmaceutical mar- ket. It developed, manufactured, and sold a wide range of products for the diagnosis and treatment of ophthalmic disorders; that is, those used in the treat- ment of defects and diseases of the eye. The company marketed both prescription and nonprescription drugs, a wide variety of products for use in ophthalmic 114 surgery and for the care of hard, soft, and gas-perme- able lenses, and a few dermatological products. Despite continuing price pressure from the cheaper generic drugs entering the market, EyeOn revenues and profits nearly tripled over the period between 2005 and 2015. In 2016, EyeOn sold products in over 80 countries, and worldwide sales totaled almost $1 billion. EyeOn used a traditional product-line organization structure. Reporting to F.ye0n's CF20 were the manag- ers of each of the product lines, and the managers of Science and 'l'echnology (Mr. DeMartino), Finance and Administration, International, and Legal. The R&D organization RSLD was critical to the maintenance of EyeOn's rate of growth. Over 25% of 2016 sales were from products released in just the past ve years. The company's RSI-D department was headed by Dr. Prakash Kuntar, who reported to Mr. DeMartino. The department included 350 people 290 scientists (80 of whom had PhD degrees) and 60 support staff. The pur- pose of the R&D organization was to develop new, mar- ketable eye- care products that would fuel the company's growth. EyeOn's board of directors estab- lished broad research policies based on the long-term strategies of the marketing divisions, but the board depended heavily on Mr. DeMartino and Dr. Kumar to provide the guidance and direction necessary to ensure effective research activities. Mr. DeMartino and Dr. Kumar complemented each other well in terms of knowledge and experience. Mr. Del'vIartino had an in-depth knowledge of Eye On's products and markets because he had advanced through the sales organization. Dr. Kumar was a chemist. The R&D department was organized inmatrix form, with eight key senior directors [plus staff support] reporting to Dr. Kumar. On one dimension of the matrix were four medical specialty groups: ophthalmology, optical, dermatology, and basic research. Personnel .in these groups specialized in particular types of diseases. The basic research group was distinguished from the other three medical specialty groups in that its work took place early in the drug development cycle [described below). On. the other dimension of the matrix were four preclinical science departments: microbiology, chemistry, toxicology, and pharmaceuti cal sciences. Personnel .in these departments were experts in one of these scientic elds. Each research program and project was managed by a medical spe- cialty expert. The preclinical science personnel were assigned to programs and projects when needed. They often had more than one assignment at any particular time- The personnel on the re search staff had needs and characteristics that were different from those of employees in other parts of the EyeOn organization. Managers in the R&D department had to be sensitive to those differences. The senior director of the basic research group explained: We're not an organization comprised of conform- ists, and we don't want to be. The other companies can have those people. Good researchers are unique. Theyr are creative and intelligent, and although they can be aloof and seemineg disen- gaged, they will work their tails off when they get on a project they like- But their feelings are easily hurt. It's very easy to kill ideas. We have to be careful because if we use punishment, discouragement or penalties for failure, we may never get another idea. Product development cycle- The product development cycle in pharmaceutical com panies such as EyeOn was long, typically totaling up to 15 years for a totally new drug and from_ 3 to 5 years for a simple product. Often the cycle started with some basic research designed to provide a better 11nd erstand ing of the underlying biochemistry of the disease pro cesses at the molecular level. In 2016, EyeOn had five basic research programs underway, all in the area of ophthalmology: inammation, immunology, glaucoma, diabetic retinopathy/cataract s, and drug delivery. When a new product concept was formed, the prod- uct development cycle was said to begin. Development consisted of a number of relatively distinct steps. First was the discovery phase of development, the purpose of which was to identify compounds with potential EyeOn Pharmaceuticals, Inc. commercial applications. Scientists designed and tested new drug compounds against the characteristics of the diseases they were studying both in test tubes and later in live animal subjects. For most new drug concepts, these screening and testing activities would last from two to ve years. When the compounds moved into the discovery phase of development, EyeOn management assigned the effort a development program number. This num ber identied the effort until the product entered the clinical phase of testing. In 2016, EyeOn had a total of 11 development programs underway, 4 each in oph thalmology and optical, and 3 in dermatology. A successful culmination of the discovery phase of development was marked by the identification of a compound that showed promise. Such compounds were moved into the optimization phase of develop ment. This phase usually involved one to two years of studies of how the compound might act in the body. Scientists would study how the compound was absorbed, distributed, metabolized and excreted in animal subjects. They would do some exploratory test ing of toxicity (i.e. harmful side effects) and stability (Le. length of time the drug retains its effectiveness when stored). By the end of this phase of development, the scientists would prescribe a preliminary chemical formulation and make a preliminary packaging de ci sion (i.e. mode of delivery and size of dosage). Drugs continuing to show promise were moved into the preclinical phase of development. This phase involved better controlled laboratory experiments to validate the results of the exploratory tests conducted in the optimization ph ase of development. The preclin ical phase of development usually lasted about 612 months. The drugs that continued to show promise were led as IND [Investigation of a New Drug) candi dates with the US Food and Drug Administration (FDA). At this point, a reasonably complete compo si- tion and specication existed, and a manufacturing procedure suitable for the prep aration of clinical sup plies was .in place. Once the IND was led, the project moved into the clinical phase of development. This phase involved tox icity and stability testing of a lon gerterm nature than had been done previously. The testing was performed on live subjects: r st normal human subjects and then dis-eased human subjects. During this testing, the sci entists would malte judgments of the safety and ef cacy of the drug candidates and make nal decisions about the dosages and modes of delivery to be used. 115 Chapter 3 - Action, Personnel, and Cultural Controls The clinical phase of development generally lasted between ve and eight years. When a product entered the clinical phase of development, a project number was assigned. This number would stay with the effort until the product received FDA approval or the effort was abandoned. In 2016, EyeOn had a total of 30 active projects. A drug that passed clinical testing was led as an NBA (New Drug Application) with the FDA. The FDA approval process took from one to three years. Approval was needed before the drug could be mar- keted in the United States. However, the product could be sold in many other countries after it had passed clinical testing. Exhibit 1 shows an overview of the product develop- ment cycle. The times shown in the exhibit for comple- tion of each of the phases in the cycle are for development of major drugs. For fairly simple drugs and optical devices, the times were considerably shorter where INDs were often led within 12 months, and clinical testing took between 12 and 18 months. A shift in emphasis Through most of its history, EyeOn had relied heavily on other pharmaceutical companies not involved in ophthalmic markets as sources of new product ideas. EyeOn scientists would screen compounds developed from these companies, and if they showed promise, EyeOn would license the compounds and introduce tai- lored forms of them into ophthalmic markets. Com- pounds screened in such a manner were entered into the product development process in the preclinical phase of development because the properties of the compounds were already understood. In recent years, however, EyeOn management had been shifting their RSzD efforts to emphasize more basic research. As the ophthalmic markets had grown, other pharmaceutical companies had entered some of EyeOn's market segments. These companies were less prone to offer EyeOn their newest compounds. Thus, the research focus had been evolving toward larger- scale, longer-term studies of more complex and sophis- ticated diseases of the eye. This is because EyeOn already had a broad product line covering most niches in the eye-care market, and to meet the company's aggressive growth targets, new breakthrough products were needed. The inevitable shift toward more basic research made management even more concerned about having measures of research productivity 116 available for control purposes because the investments in basic research were longer-term and riskier. As Mr. DeMartino explained: What's important in conducting research is to keep achieving progress on a daily basis. When it takes ten years to develop a product, you can't wait until tomorrow to get the work done. The important questions are: Are we doing everything we can to ensure that we are being productive every day? And how can I tell if we're being productive? Drug investments and payoffs New product development involved high-risk invest- ments for potentially lucrative payoffs. Across the industry, only about 1 of every 10,000 compounds investigated in the early exploratory research stages eventually proved to be commercially successful. The probabilities of failure of a typical compound in each of the phases of the product development cycle were approximately as follows: Discovery Preclin ical Negligible FDA 3. Patent The payoffs from the research were highly depend- ent on the magnitude and duration of the competitive advantage that EyeOn enjoyed when the new pro ducts were developed. Some drugs were breakthrough prod- ucts that provided significant advantages over the com- petition in large market segments. Others were either minor modications of already-existing EyeOn prod- ucts or were aimed at small market segments. Some- times competing firms developed alternatives to commercially successful new drugs in periods as short as two to three years, while on other occasions EyeOn pro ducts were sold for 20 years or more with little or no competition. As Mr. DeMartino noted: From my perspective, it's not very important whether a product costs $30 million or $60 million to develop. When we are working on a drug that will give us several billion dollars in sales over 15 years and a 75 percent gross margin, overspending a little on research doesn't matter much as long as the drug gets created. The timing of the development efforts was critical. If the development of a particular drug was pursued too early, the company could be subject to a high probabil- ity of failure and/or significant extra development expenses and, if problems were found after introduc- tion, possible legal liability expenses. If the develop- ment was pursued too late, the result would be a \"me-too\" product. Planning and budgeting EyeOn used a well-developed set of management sys- tems to help manage its R8zD effort. Planning and budgeting was done on an annual cycle, which took place from mid-July to mid-September. Planning was an iterative process. Mr. DeMartino and Dr. Kumar began the process by setting program and project objectives and priorities and by outlining an overall budget for the R&D department. In establishing these guidelines, they met with EyeOn directors and top- level managers to ensure that they had a good under- standing of market trends and the amount of resources the company was willing to spend on R&D. Then direc- tors and managers in each medical specialties group and each preclinical science department determined the labor hours and resources required to satisfy pro- ject and program objectives. This process was accom- plished through a series of meetings between directors and managers. As compared to plans for the development projects, plans for the basic research programs were easier to prepare because they used few resources from the pre- clinical science departments. Thus, very little cross- organizational coordination was required. Most development projects required the assistance of all, or at least most, of the preclinical science groups, so many meetings between the managers of the medical spe- cialty groups and the director of Pre-Clinical Sciences were required to ensure that resources were allocated appropriately and, if necessary, that steps were taken to procure additional resources. After the plans were prepared, Dr. Kumar reviewed them and made suggestions and adjustments as neces- sary. Then the plans were consolidated and compared with the overall targets. Sometimes further adjust- ments were necessary. EyeOn Pharmaceuticals, Inc. By February, all EyeOn employees were required to develop, in consultation with their immediate supervisor, personal objectives for the year. The company did not require the use of a standardized form or format for documenting these objectives, but the objectives had to be written down, and this docu- ment had to be signed by both the employee and the supervisor. During the year, budget updates were prepared on a quarterly basis consistent with the planning schedule. The budget analysis process, like the annual planning processes, was very informal because, as Dr. Kumar explained: We do not expect the scientists to act like busi- nessmen when they plan new product activity. We want to encourage them to develop new ideas with- out many constraints, and they don't like a lot of paperwork. lndeed, one research program manager explained his dislike for paperwork requirements: We work only on programs with payoffs so poten- tially large that a monkey can run the figures show- ing that the investment is worthwhile. The trick is to make the new product work, not to try to figure out that a new breakthrough therapy for glaucoma will pay off. It will! Mr. DeMartino had two main concerns about the planning proce ss. First, he wondered if too much detail was still being required. Second, he wondered if requir- ing numbers about the research activities made the managers and scientists conservative in presenting their ideas. Given the company's need for good ideas, he thought it was important that no administrative bar- riers to ideas be erected. Measurement and reporting Accounting in the R&D department was done on a full absorption cost basis. Direct expenses, both labor and materials, were charged to specic programs and pro- jects. Labor was charged on the basis of time sheets completed weekly by RSzD personnel. Costs not speci- cally identiable with a particular project or program were allocated monthly on the basis of direct labor hours. EyeOn produced an extensive set of cost reports. Many of the reports were on a project, program, or medical specialtybasis. They showed costs by line item 117 Chapter 3 - Action, Personnel, and Cultural Controls compared to budget. These reports were available on a monthly basis. Another set of reports showed expenses aggregated by cost center. The R&D department was divided into 75 cost centers. The cost reports were sum- marized by type of medical specialty and by type of project or program. The program/project cost account- ing system provided the information necessary to mon- itor the flow of resources to medical specialty areas, research versus development, and long term versus short term. The proj ect/program and cost center reports were sent to the managers responsible for the costs. The managers reviewed the reports, but they were not required to explain variances. This was because most ofthe variances were caused by changes in the scope or timing of the project/program, and such changes were almost always preapproved by Mr. DeMartino and/or Dr. Kumar. EyeOn management recognized that the cost reports were useful for measuring the inputs to the R&D processes, but they were not useful for measur- ing the productivity of the R&D activity because they did not reect any outputs. The outputs, which might be measured in terms of prots generated or value created, would not be known for years. The signi- cant lag between the investments in R&D and the returns generated ensured that traditional account- ing measures, such as return on investment, were not very meaningful except in very long measurement windows. To date, Mr. DeMartino and Dr. Kumar had focused their attention on the department's consolidated nan- cial summary (actual versus plan) and on the major R&D achievements of the year. In the last few years, these achievements were as follows: INDs led Research publications Patents indicated allowable They realized, however, that none of these indica- tors was a totally reliable indicator of forthcoming com- mercial success. 118 Incentive plans EyeOn used two formal incentive plans offering cash awards for good performance, one for scientists and one for senior-level managers. The scientist incentive plan was introduced in the R&D department in 2013. Four cash awards of $25,000 each were made annually for technical excellence. The awards were split between scientists doing basic research and those involved in development activities. Candidates for the scientist award were nominated by senior directors in the R&D department. The candi- dates' accomplishments were judged by a seven-person committee which included four working-level scien- tists, two director-level managers, and one person from outside R&D (e.g. from corporate marketing). The com- mittee assigned the awards based on \"perceptible con- tributions or unusual problem-solving capabilities which are perceptible to fellow workers.\" The management incentives were provided through a company-wide program which provided stock options and bonuses to managers down to the director level of the firm. Each year an incentive award pool was assigned to the R&D department based on a predeter- mined percentage of EyeOn prots. This pool was allo- cated by R&D management to R&D employees included in the plan in conjunction with the annual performance review. For purposes of assigning the awards in the R&D department, R&D employees were classified into three categories of achievement: (1) distinguished performance (DP), (2) superior performance (SP), and (3) good solid performance (GSP). (A fourth cate- gory called \"Needs Improvement\" was also used on occasion, but, as Dr. Kumar observed, \"these col- leagues don't get to stay very long.\") Table 1 shows the approximate percentage of pe ople who were classied in each category of achievement and the bonuses that could be expected in an average year in each of the categories. Table 1 The evaluations were based on a weighted average of three factors: (1) meeting the technical milestones in the annual plan; (2) discovering new product candi- dates; and (3) getting new products with commercial potential through FDA approvals. The factors used for weighting accomplishments in each of these areas were pre-established at the beginning of the year. In general, the highest weightings were given to the accomplish- ments that could be measured in a tangible fashion in the next 12 months. The weighting factors varied signicantly among the various areas of the department. For example, man- agers in development areas (as opposed to those in basic research) were expected to have products pro- gress through the FDA approvals, but they were not expected to generate many new product leads. The standards used to assess performance also varied signicantly among the areas, reecting the probability of payoffs of the various activities. For example, to be evaluated as SP (superior perfor- mance}, managers of basic research activities might be expected to achieve 40% of their objectives in a given year. For managers of ophthalmology drug development activities, however, the achievement of 50% of their objectives might qualify only as GSP; SP might require the achievement of 70%. For product development managers in optical, GSP might require achievement of 60% of their objectives, and SP might re quire achievement of 80%. Management concerns EyeOn managers felt they had an excellent research team that had produced an ever-increasing set of new products that had fueled the company's growth. Mr. DeMartino highlighted a critical success factor hiring: Eighty percent of the really good ideas those that lead to breakthrough products come from 20 per- cent of our colleagues. It is important for us to hire as many of those good people as we can, and per- haps even more important not to lose any we've already employed. But it is very difcult to tell who EyeOn Pharmaceuticals, Inc. the really good people are until their accomplish- ments are apparent, and that might not be for some years after they were hired. Mr. DeMartino's continuing concern was that EyeOn did not have a good early warning system in place to signal potential problems on a timely basis because of the difculty in measuring RSzD productivity, and this might be particularly costly as the emphasis shifted toward more basic research. Mr. Kumar had two related concerns. One was the challenge he faced in demonstrating the productivity of his department to his boss and the board of directors. In defining what we mean by productivity, we have to be careful in how we define our terms and the measures that we rely on. For example, we rarely terminate projects, but we do adjust priorities and let some of them sit in an inactive state until a solu- tion to a particular problem surfaces. Should the inactive projects reflect negatively on our produc- tivity? . . . It's important that whatever measures we use be simple enough to assemble and use without devoting too much time away from the job at hand doing promising, leading research. His second major concern was about the growing com- plexity in his department: We now have 11 programs and 30 projects under- way, and the growth has made coordination of the groups more difcult. It is increasingly difcult to keep up with the status of each program and pro- ject well enough to be able to decide priority issues. In the last six months, we have started an effort to try and identify a set of standard product development milestones and decision points around which a computerized information system could be built and used for control purposes. Because of the great variance among projects, however, not everybody is convinced how much use there is to organize an information system around a standardized process that doesn't really match any real project. 120 Exhibit 1 Product development cycle Phase of Basic Clinical FDA Research Discovery Optimization Preclinical development Clinical Tasks Exploratory Synthesis Preliminary Formulation FDA Review Testing Studies Material Formulation Laboratory Characterization Exploratory Testing Screening Testing Studies Typical Slapsed Several 2-5 Years 1-2 Years 6 months- 3 months- 2-3 Years 2-4 Years 1-3 Years Time for Major Years 1 Year 1 Year Systemic Drug Major New Drug Lead IND IND Safety Efficacy NDA NDA Milestones/ Concept compound Candidate Filed Viability Viability Filed Approved Decision Points Identification Decision Decision Decision Drug Preliminary Final International Domestic Performance Packaging Dosage Product Product Criteria Decision Decision Release Release Final Licensing Process Agreement Decision