Elsevier

Drug Discovery Today

Volume 13, Issues 21–22, November 2008, Pages 997-1001
Drug Discovery Today

Review
Post Screen
Drug discovery: are productivity metrics inhibiting motivation and creativity?

https://doi.org/10.1016/j.drudis.2008.06.015Get rights and content

With a productivity gap in pharmaceutical research and development, and increased industrialization in both areas, an increased need for precise indicators of productivity has emerged. Measuring scientists’ performance can impact the way the tasks are performed and the level of motivation of an individual. This is a crucial aspect when key performance indices of other performance metrics are to be defined within an organization. Motivation is a main driver of creativity and should, therefore, not be compromized by a need to measure productivity. This paper is based on 120 interviews in over 50 companies from 2005 to 2008. The results suggest that the level of detail at which performance should be measured depends on the level of industrialization that a technology falls within. Performance metrics are a means for feedback to individuals. Furthermore, we show that the level of motivation is not directly correlated to the level of detail that a group's performance is measured at, but instead that it varies from person to person. Consequently, we suggest that the level of detail of performance measurement and the motivation profile of the scientists need to be aligned. This is an important aspect to consider when measuring performance.

Introduction

Sources of innovation differ depending on the stage of maturity of the organization. As highlighted by Ullman and Boutellier [1], larger research organizations go hand in hand with a shift from personal initiative to a need for coordination. Quinn and Cameron [2] described the four stages of this evolution as (a) the entrepreneurial stage; (b) the collectivity stage; (c) the formalization stage and (d) the elaboration stage.

According to Quinn and Cameron, these stages imply different sources of innovation from (a) the owner; (b) employees and managers; (c) a separate innovation group and (d) an institutionalized R&D department.

In large organizations with a high need for productivity in research, institutionalization seems to be only partially beneficial for innovation.

Because of increasing merger and acquisition activities – integrating biotechnology firms into larger research organizations and also within large global research organizations – various cultures and departments belonging to different stages of the evolution described above need to coexist. Thus, the entrepreneurial stage, at the laboratory level, has to cohabit with the formalization and elaboration stages and their respective sources of innovation at the departmental level.

Drug discovery involves a large number of specialists in diverse areas engaged with numerous scientific communities and technologies. These technologies are in various stages of maturity. March [3] refers to an evolution of technology maturity from exploration to exploitation. Abernathy and Utterback [4] describe the technology life-cycle as an evolution from product innovation over dominant design to process innovation (Fig. 1) in three phases: (a) the fluid phase, where extensive experimentation with product design occurs; (b) the transitional phase, where process innovation prevails over product innovation and (c) the specific phase, where product and process innovation decrease and emphasis is set on cost reduction.

Monoclonal antibodies (mAbs), interference RNA (RNAi) and genomics or proteomics belonged to the fluid phase in the 1980s. Now, according to our interviews, RNAi and mAbs have emerged as dominant designs, where extensive process innovation occurs. This is the transitional phase. Finally, in vivo drug metabolism and pharmacokinetics (DM–PK) assays could be classified in the specific phase, where product and process innovation have slowed down.

The fluid phase is often performed in universities and biotechnology firms where high risk is tolerated. The pharmaceutical industry picks the technologies that then often become a dominant design, for example, combinatorial chemistry or HTS. Technologies that have reached the specific phase are candidates to be outsourced to countries with lower labor cost, because the task can easily be specified and quality can be measured [5].

With this evolution into more process-driven activities emerges an increased need to measure processes more precisely, to gain efficiency. The reason is that, with each point of measurement, one can limit the number of possible sources of waste, akin to localizing a leakage in a pipe. With more detailed measurements of performance arises a challenge to keep researchers motivated and creative. This mix of innovation sources and levels of bureaucracy and industrialization makes it even more challenging to measure performance, especially in large organizations. Furthermore, Ullman and Boutellier [1] highlight differences in performance measurement between different laboratories at the same hierarchical level, depending on whether the activity is creativity- or process-driven. Accordingly, within the hierarchy of an organization, performance measurement varies both vertically and horizontally.

Measuring performance can affect the investigated individual's actions. Management goals are set by metrics or key performance indices (KPIs) that should contribute to individual, group and company strategic goals. Depending on the activity, it is more or less easy to find an adequate metric. According to Roberts’ [6] ‘simple agent theory’ [7], performance metrics need to be informative, reflecting the individual's choice of effort. Metrics influence what an individual or a group does and how they accomplish a task. Hence, with metrics, management has an impact on individuals’ delivered performances. Motivation is the sine qua non for creativity [8] and initiative. Thus, impact on motivation needs to be considered when measuring performance, assuming that creativity and individual initiative is beneficial to the firm. The performance of some activities can be measured precisely, whereas the performance of others can only be determined vaguely. For example, the activity of a high-jumper compared with that of an ice skater is instructive. The performance of a high-jumper is measured by the height he or she can jump. The ice skater is asked to make a triple spin, which involves speed, a number of rotations, staying centered, keeping in control and elegance. The first four markers of a successful triple spin can be measured in absolute terms, but elegance is a subtle variable assessed by experts, based on their experience and preference. Similarly in science some activities are predictable and measurable in absolute terms, whereas others can be assessed only by experts [1].

We define the ‘detail of performance measurement’ as the degree to which a process is predefined by the measurement. Consequently, if the detail of a performance metric is high, there is limited space for an individual's own initiative on how to perform the expected task. According to Hackman and Oldham [9], autonomy in configuring how a task should be carried out is an important component to motivation.

Measuring performance is often regarded as an instrument panel for managers, although its primary aim should be to give objective feedback to (and within) the organization. Feedback is also a major driver for motivation [9]. If not measured at all, an individual feels that the task he or she is doing is not important. Feedback is a means for recognition, which is a major driver for many scientists. Similarly, when playing football people are more motivated when the score is counted, as opposed to if the result does not count. If performance is not measured, there is no benefit to work hard. At the other extreme, if performance is measured at a highly detailed level micromanagement reigns and employees begin to feel controlled and resentful of a lack of autonomy.

We therefore assume that somewhere between these two extremes exists an optimum – a detail of performance measurement that we will call “performance-driven empowerment”, where an individual's motivation is at its highest (Fig. 1).

We assume that this maximum depends on a person's intrinsic motivation and, thus, cannot be influenced by external incentives [8]. Our hypothesis is that this optimum differs from one person to another (Fig. 2).

Section snippets

Method

We interviewed scientists in over 50 companies that performed numerous activities involved in the drug discovery process. In total, we carried out 120 interviews in GPCs, CROs and biotechnology firms in China, India, USA and Europe. A case study method [10] was used. We asked (a) how their performance was measured, (b) if they felt this was a fair assessment, (c) if they felt they were limited in their personal initiative due to a restrictive performance system, (d) if the system gave them the

Results

In our sample of companies we noticed different ratios of PhDs to MScs within groups, depending on the activity performed. Given high levels of industrialization within a group, specialists were able to predict and plan a process in a rigorous way. Thus, prospective criteria were defined as quality properties. There was consequently less need for expert assessment when measuring performances, and the ratio of PhD to MSc scientists within the team decreased with the level of industrialization.

Concluding remarks

To keep motivation high within research groups, management needs to consider each individual scientist based on his or her need for detailed and frequent feedback. Each activity needs to be staffed with scientists whose motivation curves reflect the level of industrialization of the activity they are responsible for. Because the level of industrialization changes over time, restaffing is required to keep a high efficiency within the research group and to maintain a balanced cost structure. The

References (10)

There are more references available in the full text version of this article.

Cited by (13)

  • What is the most important approach in current drug discovery: Doing the right things or doing things right?

    2012, Drug Discovery Today
    Citation Excerpt :

    An individual's objectives should always be aligned to the objectives of the team and the team's objectives should be clearly aligned to the drug project's objectives – with the project's objectives of course being aligned to the company objectives. Clearly, fostering an environment focused on delivering a robust and sustainable longer-term pipeline through high-quality innovative science is paramount and, as such, we must be mindful not to compromise scientists’ motivation by too detailed, short-term focused performance metrics (for a discussion of how productivity metrics can affect motivation and creativity, see Ref. [12]). Leadership and management are both very important.

  • Translational semantics and infrastructure: Asearch for the emperor's new clothes?

    2012, Drug Discovery Today
    Citation Excerpt :

    Aside from the viewpoint that existing therapeutics produce their effects at the ‘easy targets’ or ‘low hanging fruit’, and that newer targets are exponentially harder to ‘drug’ [65], three additional factors have contributed to the cost and/or success mismatch. The first is an overt focus on metrics [66], driven by computer-based technologies and used for: (i) data mining to predict relevant targets and compound efficacy; and (ii) designing, synthesizing and assessing the activity of compounds [67]. This has resulted in more compounds being synthesized and analyzed over the past decade than in the history of biomedical research.

  • To measure is to know: An approach to CADD performance metrics

    2011, Drug Discovery Today
    Citation Excerpt :

    The metrics themselves consider only the outcome of the work rather than the process to achieve it. We believe, and so far have observed, that such an approach to metrics will largely avoid the stifling effect on innovation that is often attributed to them [9]. Our approach to metrics collection is largely automated by leveraging pre-existing databases that senior management has already mandated.

View all citing articles on Scopus
View full text