Management of R&D - Reaction Paper No. 7

1. Motivating and Measuring R&D Productivity

A. “Categorizing the Measures and Evaluation Methods of R&D Performance”

(Summary) To measure R&D performance effectively, it is necessary to consider the many dimensions of performance analysis and understand how they influence various R&D performance measures or evaluation methods. Among the dimensions are as follows: 1) measurement perspectives, 2) purpose of measurement, 3) level of measurement, 3) type of R&D, and 4) process phase. Examples of measurement perspectives are customer perspective, internal perspective, financial or stakeholder perspective and learning perspective. Measurement perspectives are areas of measurement derived from a firm's strategy and strategic objectives. In essence, this dimension is about the diverse point-of-views from where measurements can be taken. Different measures or metrics are associated and can be categorized according to these perspectives. Applicable or suitable measures can also be categorized according to their purpose which can be for motivating and rewarding people, or diagnosing activities and organizational units. The latter can further be sub-divided into more specific purposes like strategic control, justification, benchmarking and resource allocation. Level of measurement also influence which measures are applicable. The possible levels of measurements are national, industry, network, company, strategic business unit / department, R&D process, R&D project, R&D team and individual researcher level. The type of R&D, which are basic research, exploratory research, applied research, development and product improvement, also dictate the type of suitable measures. For instance, qualitative measures are best suited to basic research while quantitative measures are best suited to product improvement. The process phase is another dimension that needs to be considered when selecting measures. The different phases correspond to different subjects of measurements. The different phases are inputs (people, information, ideas, equipments, etc.), in-process (R&D lab, receiving systems, etc.), outputs (publications, new products or processes, patents, etc.) and outcomes (cost reductions, sales, product improvements, etc.). When selecting R&D performance measures, it is important to both recognize the combinations of dimensions that must be taken into consideration and be familiar with the various applicable measures associated to each dimension. Furthermore, it is crucial to understand both the relationship between the dimensions and the measures and the intersection of these relationships.

(Reaction) For R&D performance measures to serve their purpose, which is to provide an accurate picture of R&D performance, they need to be selected through a thorough and thoughtful process. The context and purpose of the measurement needs to be thoroughly considered and understood before specific measures can be selected because only by doing so can the appropriateness of any measure be reasonably judged. In other words, we need to know what it is we are trying to measure and why before we can know how to 1) measure it effectively and 2) use the measurements productively.

B. “Innovation and Incentives: Evidence from Corporate R&D”

(Summary) A study conducted on 140 publicly-traded firms to understand the relationship between incentives given to corporate research executives and innovation found that among firms with centralized R&D, there is a clear positive correlation between long-term incentives, such as grants of restricted stocks and stocks options, and the following innovation measures: patent awards, citations and originality. However, there was no correlation found between short-term incentives and innovation. Along with compensation data for corporate R&D heads, the study also collected compensation data for the other senior management positions (i.e. chief executive officer, chief operations officer, chief finance officer, human resource head) to serve as control. The compensation components collected include executive salary, actual bonus, grants of restricted stocks, stocks options and other long-term compensation units in a given a year. Innovation data collected include the number of patent applications and awards to the firm in a given year, the mean number of citations to the firm's patents in a given year and the mean “originality” of the firm's patents in a given year. (A patent's originality is measured in terms of the number of citation made by the patent to earlier patents. Patents that draw on a narrow array of patents are considered more original.) The subject firms were chosen across industry sectors. Results of the study have shown that the mentioned correlation exists only among firms with centralized R&D but not on firms with decentralized R&D i.e. firms that reported divisional R&D heads as well as corporate R&D heads. The correlation (between long-term incentives and innovation) itself is probably due the high quality of decisions made by R&D heads, especially during project selection, that leads to more productive R&D efforts. Its existence in firms with centralized R&D only can be attributed to the extent of influence and authority that R&D heads in these firms posses over R&D direction and activities, which is lacking in firms with decentralized R&D. The study did not investigate the relationship between incentives given divisional R&D heads and innovation. It also did not investigate relationship between incentives given to rank and file researchers and innovation. Furthermore, it cannot distinguish whether the stronger incentives lead to R&D managers making better decisions or lead to attracting more innovative R&D managers that made better decisions.

(Reaction) The study provides an interesting (though limited) insight into the link between long-term incentives and innovation. However, it does not address the issue that perhaps the measures of innovation that applies to organizations with centralized R&D and organizations with decentralized R&D may be different and that perhaps, the measures chosen in the study favors those with centralized R&D. For instance, since centralized R&D are more likely to work on fundamental research (i.e. basic research) than decentralized R&D, they are more likely to produce more general results i.e. results that are applicable in many areas and therefore, their patents would tend to get more citations (since they are more general as opposed to typical results produced by decentralized or divisional R&D which are specialized and specific).

2. R&D Project Management Techniques

A. “Knowledge Management Practices for Innovation: An Audit Tool for Improvement”

(Summary) Knowledge management practices (KMP) are activities within the firm that drives or supports development and application of knowledge. They provide a powerful means of understanding and relating knowledge management and the innovation process with the goal of improving business and R&D performance. KMPs are valuable because: 1) they provide a tangible framework to analyze knowledge management within innovation processes., 2) they have common features across companies that allows for best practices to be identified and transferred, 3) they allow for regular audits that help firms broaden and continuously improve their practices, and 4) they enable companies to radically depart from ineffective cultural and technical traditions. KMPs for innovation are the practices that directly contribute to the creation of novel business propositions and are primarily concerned with accumulation, analysis, management and dissemination of stocks of knowledge in a firm, covering three main knowledge areas: technology, markets and company processes. These KMPs can be categorized into five groups which are as follows: 1) those relating to R&D management process, 2) those relating to mapping of knowledge relationships, 3) those relating to human resource management, 4) those relating to management of intellectual property positions, and 5) those relating to R&D information technology management. The 'R&D management process' group is concerned with building, maintaining, sharing and managing of knowledge and experience gained from R&D activities embodied in the R&D scientists and their formal and informal communication patterns within and beyond the lab. The 'mapping of knowledge relationships' group is concerned with coordination of internal and external R&D capabilities, inter-firm relationship and market requirements that spans disciplinary, organizational and company boundaries with the use of capability maps to identify locations within the company of formal knowledge bases, skills and supporting equipment and services. The 'human resource management' group is concerned with motivating and rewarding R&D personnel to encourage knowledge sharing and development of inter-disciplinary expertise. The 'managing IP positions' group is concerned with the legal implications of IP as well as leveraging IP to update and manage corporate stocks of knowledge to improve business and R&D performance. The 'R&D information technology management' is concerned with applying ICT advances to transform information management practices: 1) to make knowledge more accessible, 2) to facilitate knowledge diffusion, 3) to allow for more effective use of stocks of knowledge, and 4) to trigger to create or change KMPs in the other groups. For instance, it could involve establishing shared ICT systems and interfaces across organizational structures, possibly including intended customers. Based on these groups, an audit tool in the form of a questionnaire and an action plan form was developed to help companies highlight the importance of each KMP to the organization and identify areas of KMP where improvements could have the optimum impact. Furthermore, it enables a company to develop its own KMP profile and KM strategy and understand the relationships between knowledge management and the innovation process in order to help sustain long-term business success.

(Reaction) In the context of KMPs, there is tremendous potential waiting to be unlocked in using collaboration tools such as wikis and content management systems. When implemented and turned into practice throughout the organization, it does not only facilitate knowledge sharing and conversion of tacit into codified knowledge but also allow individuals leverage other people’s knowledge with very little overhead. Furthermore, it has a synergistic, multiplier effect. That is, the more people use it, the more useful and valuable it becomes (i.e. it follows Metcalfe’s law).

B. “Project Management Theory and the Management of Research Projects”

(Summary) A research project manager is responsible for supporting creative thinking in small subject-oriented units and ensuring it results to a concrete output in the form of codified new knowledge or concrete technologies or technological processes within budget and time constraint. In short, he manages knowledge and knowledge workers and this means he must deal with complexities arising from the nature of researchers and researcher work and the uncertainties associated with generating research results. This complexities include the following conflicting forces: 1) researcher's desire for autonomy and the need for strict project control, 2) cooperation and competition between researchers in the project, 3) need for predictability of project output and unpredictability of research outcome, 4) knowledge asymmetry between project manager and individual researcher, and 5) the need to take risk to innovate and the need to reduce risk to ensure delivery of desired results. To successfully deal with these issues, he must: 1) facilitate open and effective communication, 2) manage conflicts, 3) present a unifying vision and nurture a project environment that turns a group of individuals into a committed, empowered, self-managing team which functions as a single integrated, collective mind i.e. “the distributed mind”, and 4) be a “boundary manager” who manages channels of influence and external relations and is able to  shield his team from outside distractions and confusion. In short, he must both be a leader that focuses on people and on getting the right things done and a manager that focuses on optimizing processes and getting things done within time and cost limits. Management of research projects differs from management of manufacturing or construction (upon which most PM literature are based) in that research projects are dynamic, uncertain and requires continuous learning, adjustment, adaptation which is why higher priority needs to be given to the human processes – the soft side of project management - and not just focus on the technical structure aspects – the hard side – such as the tools of planning, scheduling and controlling.

(Reaction) It is striking how the concepts and ideas, like having small, empowered, cross-functional, self-managing teams, presented in the article in response and to address uncertainly are precisely the same ideas promoted by the relatively new agile software development methodologies (such Scrum, XP, etc.) that emphasize the value of collaboration and iteration.