ارزیابی پرتفولیو تحقیق و توسعه پایدار

Research and development portfolio management is traditionally technologically and financially dominated, with little or no attention to the sustainable focus, which represents the triple bottom line: not only financial (and technical) issues but also human and environmental values. This is mainly due to the lack of quantified and reliable data on the human aspects of product/service development: usability, ecology, ethics, product experience, perceived quality and the like. Even if these data are available, consistent decision support tools are not ready available. Based on the findings from an industry review, a DEA model has been developed that permits to support strategic R&D portfolio management. The usability of this approach is underscored with real life examples from two different industries: consumables and materials manufacturing (polymers).

1.1. Description of the problem context Today's innovation strength is more than ever determined by a company's ability to differentiate from the competition in delivering customer delight at the pace of the market, applying the best available technology. Sustainable innovation results in profitable, people oriented, and planet-minded products and services. It is a great challenge for innovation and R&D managers to permanently assure that R&D budgets are allocated to the best set of R&D projects in order to reach the innovation targets on both the shorter and the longer term (Cooper [10], Cooper et al. [11]). Factors with a traditionally strong influence on strategic innovation decision making, such as the business opportunity and the feasibility of a project, do not necessarily predict project success (Moenaert et al. [25]). From their study, competitiveness reveals to be a strong predictor for success. Competitiveness consists of the following three elements: a competitive answer to a threat or an opportunity, the size of the advantage over the competition (incremental or game changing) and the sustainability of the innovation. Competitiveness is linked with differentiation and added value as perceived by the customer. In order to create a sustainable advantage there is a need for non-imitable features, such as non-technological, intangible aspects leading to product/service experience and meaning, in line with the brand experience and corporate identity (Borja de Mozota [5]). This holds both for incremental as well as for radical innovations. However, the latter cannot be derived from actual user and market research, since they create new markets and envision the user of the future in a future context (Verganti [37]). Radical innovations, based on new technology, or addressing new user needs or creating new markets, are hard to evaluate, especially in the early phases of product development. There are many uncertainties and risks on all three types of innovation aspects: technology, economies and values, as shown in Fig. 1. “Values” are defined as the ethical, societal and personal values and perceptions leading to a product experience. Radical and incremental innovations differ dramatically in the availability of information on opportunities and risks. They are difficult to compare and equally serve the goals set by the business and innovation strategy. Industry practice shows that a separate approach in budget allocation and R&D organization for radical innovations is usually applied. This, however, does not solve the problem of evaluating the radical innovation projects. Optimal R&D budget allocation according to the innovation strategy requires incremental and radical innovations to be evaluated simultaneously.Sustainability, as defined by the Brundtland Commission (UN documents [36]), was translated into the triple bottom line (people, planet and profit), and adopted by many companies through their mission statement and innovation strategy. However, the societal (people) and ecological (planet) dimension of sustainability are difficult to incorporate in formal decision making policies and decision support systems. From a short term perspective, designing for user experience can lead to differentiation and innovation success and thus economic benefits. When embedded into a long term company strategy, aiming for “customer delight” involves human-centered design and experience innovation and leads to both incrementally and radically new products and services addressing today's and tomorrow's user needs. However, the economic benefits resulting from “user experience” are hard to estimate in the early phases of New Product Development (NPD). Moreover, the amount in which an innovation contributes to a user experience represents a completely different type of benefit for the company than the direct and short term economic consequences. Ecology and human (or user) related aspects are part of the earlier defined as values, since they both involve ethical values, which can be commonly shared by a group of customers or they can be individually perceived. Values cannot be translated into monetary figures without loss of information. They can be very different, even conflicting in their nature and value. For instance an innovation idea can lead to a very high user experience but low ecological contribution. They will therefore be treated as separate dimensions in the decision support for strategic innovation decisions. R&D managers are in search of a consistent way to translate the innovation strategy into an R&D portfolio, taking into account all three dimensions of the sustainability concept. Available decision policies and supporting tools for strategic innovation decision making are not well suited to handle the intangible aspects of customer delight and ecology; they are too slow to respond to changes in both endogenous and exogenous factors. They favor incremental innovation versus radical innovations, especially when these originate from a user centered perspective. As a result, mainly the technological and financial (market related) aspects prevail in the ranking and selection of projects at the strategic level, defining the R&D project portfolio. The disability to evaluate a projects' performance level on the value based aspects, and to evaluate its overall performance on technological, financial and value based aspects forces ad hoc, subjective and informal decision making. This prevents a consistent value based innovation strategy and R&D portfolio. Innovation processes such as the widespread stage-gate® process, have many benefits for the management of risks and the monitoring of KPI's along the innovation funnel. However, the sequential process is not a reflection of the real-life, parallel and iterative innovation activity, and it has a tendency of favoring projects who are in their later phases of development (Repenning [29]). Next to adequate decision support models and processes, the implementation of a sustainable R&D strategy requires a new mindset, reconciling an analytic as well as an intuitive approach (Martin [24]). Decision makers will face less certainty and will have to manage more risks and must be open to creative, intangible inputs during the R&D process. This will require a behavioral change in the decision making teams. In order to gain trust amongst the decision makers, it is of utmost importance to maintain transparency and consistency in the decision making, with a strong emphasis on communication based on the visualization of the data and results. This is a cornerstone of the “Design Thinking” way of R&D management, in which a holistic approach is applied and which applies rapid visualization of an idea and validation of each concept (Lockwood [23]). The decision support system aimed for should be capable to support Design Thinking on the strategic innovation level and therefore, besides transparency in the data, the system will have to possess quick recalculation functionalities and to provide an intuitive visualization of the model results. The approach is in line with findings in the literature. For instance (Brodt 2007), discusses the observation that product success is not only dependent on the technological and economical characteristics, but also relies on the experience of the product. The latter overlaps with the third cluster as depicted in Fig. 1. Radical innovation design practice, based on product experience, has been developed by Hekkert et al. [16] through the ViP design methodology, in which user experience and contextual factors determine the innovation requirements. The corresponding dimensions are part of the “values” in Fig. 1. 1.2. Motivation: industry review The first step in this research related to decision support for sustainability in R&D portfolio assessment, is a multilayered scan on how Flemish companies managed their R&D portfolio. Several ways have been adopted to reveal the relevance of the business process named ‘R&D portfolio management’ whereof portfolio assessment is a crucial part. This field research is a consolidation from the following sources: • Systematic audits and interviews at over a hundred companies in Flanders across different sectors and ranging from SME's to large companies have given insights in the status of design management and R&D maturity. Most companies have a process and decision policies in place, but only very few realized the brand promise and company identity, including the values, in their innovation processes. • Focus groups and discussion networks with industry captains (technology and plastics sector) have been used to detect the future challenges regarding product portfolio, innovation and design management. • In-depth interviews (25) with R&D managers and CEO's in the technology sector, incl. mechatronics and biotechnology, have resulted in a detailed collection of data on innovation drivers, KPI's, portfolio management and decision processes, decision criteria, frequency of meetings, user involvement in R&D, R&D management responsibilities & constraints. From this review, it is possible to undoubtedly conclude on the real need for R&D assessment decision support in general and more specifically on the lack for a corresponding assessment model with the sustainability – and more broadly the value aspects – embedded in a decision support tool. It turns out that from a decision support point of view, the problem of R&D portfolio assessment encompasses: • A low success ratio for radical innovations mainly due to failing user acceptance, and consequently a low ratio of radical innovation investments versus investments in incremental innovations. • A priority setting gap between technology driven, and market driven R&D entities leading to long decision cycles and inefficient R&D management. • Strategic directions are directly labeled on individual projects in a straightforward way overruling the scoring models in place. These projects are hardly questioned during the course of R&D because they are imposed by another decision authority and are outside the formal decision process. • KPI's for R&D miss opportunity to drive improvement actions and target setting, because the link between these KPI's and the selection criteria is not always clear. • Sustainability seems to be difficult to integrate in the formal decision process, it is either put forward as a strategic direction and is materialized in a few strategic projects, or it is found at the end of the development process where mainly incremental improvements can be expected. • Input data based on human evaluations, used for score calculation, are not consistent. Sometimes different viewpoints, contexts or definitions are interpreted. • R&D portfolio decisions are mainly made based on economic figures: NPV, ROI and other monetary metrics. • Intangibles and value based arguments are difficult to deal with for idea selection and project ranking and are not taken into account. • Need for dynamic portfolio management tool allowing what-if analysis for changing external conditions with adequate frequency of updating. • Behavioral aspects in decision making play an important role; shortcoming in visualization impedes good communication, the interpretation of the data differs between the members of the decision making group. 1.3. Industrial practice: R&D portfolio assessment tools The industry review reveals that several decision support tools are in use today. Besides scoring models, the major characteristic of all these is that they are descriptive, leaving a huge opportunity for subjective inclusion of arguments and opinions. R&D managers state frequently that in this way consistency is hard to preserve whilst various stakeholders in the R&D assessment feel uncomfortable with the personal weight individuals can embed in the process. “It is very difficult to objectively evaluate a promising technology driven research project, of which the economic value is based on estimations, against an incremental innovation project for a product segment in a prosperous market segment. Actual sales figures or a customer's request have more weight than a future potential.” The portfolio management goals, derived from the innovation strategy, can be diverse: to maximize project efficiency, to maintain a balanced portfolio or to aim for highest success ratio. They require different types of portfolio assessment tools and visualizations. As an illustration we list a few generic tools. 1.3.1. Bubble graphs, tables and spider graphs Bubble graphs (Fig. 2) are commonly used to visualize the results checklists or scoring models. Dimensions are typically consolidated units such as “opportunity/risk/resources needed” or detailed fragmented parameters such as “customer perception/technological risk/capital invested”. In each case the number of dimensions is clearly limited to three, leaving the task of overall evaluation to the decision maker. A set of different bubble graphs are used in order to bring more clarity, but more often this leads to the opposite effect, due to the bounded rationality decision makers are subject to (Holt [17]). The graphs allow to define a “safe area” in which selected projects should be found, and they give a visual impression of balance on the dimensions chosen. As shown in Fig. 2, it is possible to emphasize strategic choices by adapting the “portfolio focus” and process indicators with rulers in the spreadsheet and thus putting more weight to the parameter. An immediate recalculation of the bubble graphs is then realized, which gives a potential for what-if analysis. The effect however, on the decision makers was confusion and lack of trust in the model. The data collection was a rigorous task, only performed once or twice a year.1.3.2. Portfolio mapping — innovation level A two-dimensional mapping gives a view of the innovation ambition reflected in the portfolio. It is used as a second order mapping tool for balance in the portfolio, returning figures for KPI's and scorecards. It is usually updated once or twice a year. Summarized, these tools have a low analytic content in the sense that they mainly try to map the data at hand. For the matter of decision support, they remain subject to personal opinions, power games, colored interpretations as well as they are limited only to a partial analysis. However, the industry survey turned out that these decision support tools are quite popular and widespread. This is mainly due to two important features: • The power of visualization. This is extremely important as these tools are used for group decision making. However, due to the two- or three dimensional restriction these tools require oversimplification of the data or consolidation of different dimensions, which makes the result less transparent. • The power of communication. This is a crucial step in order to disseminate the conclusions of the R&D assessment process. As a consequence, these two observations put a necessary condition on the newly proposed decision support tool for R&D portfolio assessment (Fig. 3).

As a matter of conclusion, the 2 research questions phrased earlier are revisited: (Q1) “Can sustainability in particular or value based criteria be integrated in general into R&D portfolio decision making?” As sustainability gets mixed with the other performance criteria, it imposes challenges on both the methodological and the practical side. From a methodological point of view, many of the value based criteria are descriptive and qualitative. Turning them into numbers is not trivial. Likert scales and categories have shown not to be that promising whilst the additional insight and value of these complicated models is questionable. Therefore, based on the two case studies revealed in this paper, it is suggested to model the value based criteria as much as possible as continuous cardinal criteria. From the practical point of view, establishing good continuous cardinal measures for value based criteria is not obvious. The quest for good metrics, which are in line with the strategic mission and options, must be intensified and appropriate data collection processes have to set up. As far as the output from the decision support system is concerned, the visualization of the results turned out to be a key success factor. This is the solid base for proper and profound insight creation which serves on its turn as the fundament for clear and concise communication. As the R&D portfolio assessment process is a typical group decision exercise, the decision support tool has to fit this setting. (Q2) “Can mathematical optimization (DEA) be used to support sustainable R&D portfolio management in a more complete, transparent and consistent way?” Data Envelopment Analysis turned out to be a suitable methodology for R&D portfolio assessment, including the value based criteria. The main advantages are the fact that it is complete (it summarizes the multiple criteria into one assessment and ranking), transparent (the assessment results can be justified and explained) and consistent (the methodology neutralizes subjectivity and circumstances). In a group decision process, it definitely takes care of the analytic part of the R&D portfolio game. Based on this analytic part, all other considerations can be added in order to come to a solid, well-thought decision. When this is done on a regular basis, opportunities in the field of R&D planning arise: project selection, ranking and clustering, resource allocation, R&D technology forecast, R&D road map construction, etc. As far as future research is concerned extensions on the proposed R&D portfolio assessment approach are to be formulated in the direction of project selection under constraints, the formulation of roadmaps for individual projects and the entire portfolio and to exploit the possibilities for portfolio follow-up and control.