Join@published by Elsevier Science Ltd, Oxford and Akaddmiai Kiad6, Budapest
QUANTITATIVE
METHODS
Scientometrics, Vol. 45, No. 3 (1999) 459-462
IN INDUSTRIAL R&D
E. VELTKAMP SeniorVice President Research Foods Unilever NL-3130 AC Vlaardingen (The Netherlands)
The R&D performance of a company is one of the key factors determining its competitiveness. The capability of companies to anticipate and adapt their approach to the dynamic trends in the society, the various markets and the industry itself is key for their success. Innovation in itself is not enough, it is the speed of innovation as well as the impact of innovation that are qualifiers for the measurement o f a company's R&D performance. Having said that the key question is how to measure the performance in a realistic and reliable way. I would like to share with you my views on that topic and before I address it let me give you some background of the company Unilever such that you have a feeling from which playing ground I make my observations and statements. Unilever is a consumer goods company mainly focused on two consumer goods businesses, namely food products (comprising 52% of the turnover) and Home and Personal care products. With a turnover of more than 90 billion guilders (figure 1997) Unilever operates all over the world with 300.000 employees involved. Approximately 2% of the turnover (last year more than 1.6 billion guilders) is invested in R&D. As you will notice from this sheet, half of the Research money is spent on central labs, whereas the other half is spent in innovation centres and development sites located in the regions close to the markets we serve. Today central R&D activities are carried out in Europe (UK and Netherlands), USA, India and China and the red marks a.,';locations of innovation centres. A number of years ago all our R&D activities were streamlined and focused in cle~." defined projects. This was not only done for development projects with a time horiz,~n of 1-2 years, but even for more exploratory projects with a time horizon o f sometimes 10 years or more. An in-house developed project management system was developed which, as a uniform IT toolkit, is applied all over the world. Representatives from business and research meet frequently at all stages of the project to determine a 'go/don't
0138-9130/99/US $15.00 Copyright 9 1999 Akaddmiai Kiad6, Budapest All rights reserved
E. VELTKAMP: QUANTITATIVE METHODS IN INDUSTRIAL R&D
go' decision and once a year the total R&D portfolio is reviewed with top management involvement. Another tool to map the total R&D portfolio for a specific business unit is the Consumer Technology (CT) matrix, where each project is mapped, based on the consumer value perception and the status of the enabling technology. The matrix provides a quick insight whether the R&D portfolio is primarily focused on incremental innovations (lower right corner) or more balanced. By making the size of the circle in proportion to the added value created by the project one also gets a view of the overall impact of the actual portfolio. Before I address quantitative methods that we use to judge our R&D performance I would like to underline a more general, but critical aspect for executing R&D within an industry: and that is the climate/culture within the company. A key question is to what extent management is committed to R&D and what is the perceived research sensitivity of the business. Is it "nice to have" or seen as an absolute need for building competitive advantage? In answering that question there are some important criteria to check: 9 Is the R&D agenda fully integrated with Business Strategy, 9 Are core technologies for the company clearly defined, 9 Clearness about what to do in-house versus out-house, 9 Transparency of the R&D portfolio, its progress, failures & successes, 9 Do progress reviews have attention of Top management, 9 Is there a balance between Exploratory activities, Research activities and Development activities, 9 Is there a structured process in monitoring and anticipating external developments. This general aspect is of critical importance because one can have perfect lab infrastructure, toolkits to measure performance, if the right climate and stimuli are not there they will add little value to the R&D process. There are a number of tools / procedures that in my view facilitate an effective R&D process within the industry. The first one is to use a clear det'med decision support system for selecting the projects. Parameters at least to build in are the strategic fit and impact, the risk analysis, the technical feasibility and obvious factors like competitive situation, costs, resources, time scales, etc.). The second one is to have, right from the beginning, business involvement in the process in order to generate commitment and ownership. Project management systems can significantly help in tracking the progress of the projects and creates a structured and disciplined way to follow the project during
460
Sr
45 (1999)
E. VELTKAMP: QUANTITATIVEMETHODS IN INDUSTRIALR&D
different phases. R&D projects, like other type of investments need these type of tools also to timely stop projects or switch the focus when the original objectives are not reacl~ed versus benchmarks determined. Project teams can be established to execute R&D projects, but sometimes teams can be frustrated since their role and accountability are not clearly defined at the beginning. Empowerment is crucial for teams to score goals. A very important aspect is to organise gatekeeping meetings when projects are moving from phase e.g., from the capability phase to the implementation and roll-out. Finally systems to evaluate the success of a project post-market launch is of enormous value also to build in learnings for future project selection criteria. Many books and conferences have already addressed how to measure R&D performance not only in a qualitative way, but also quantitative way. In my opinion there is no single criteria that can be used as an absolute barometer. In judging R&D performance we use a number of criteria such as: 9 Percentage Turnover generated by new innovations, 9 Rate of technological innovation, 9 Percentage Turnover protected by proprietary rights, 9 Patent filings / grants, 9 Technology-based diversification: spin-offs with value for other business lines (royalty flows), 9 Percentage of innovations that after market launch are a real, sustainable success (number of withdrawals), 9 Value created over time versus R&D investment, 9 Benchmarking. The importance given to each individual criteria will differ from company to company, but one lesson learned is make those criteria transparent within the organisation and communicate not only failures, but more important successes. Some final remarks: 9 Measurement tools for R&D are not a guarantee for success, 9 Avoid that tools and systems are just perceived as bureaucracy, 9 Train people adequately such that systems implemented really support and facilitate the internal processes. Last but not least: The mission for R&D as an integral part of the business system is key in terms of where R&D has to focus on, what R&D has to deliver and what value will be created.
Sr
45 (1999)
461
E. VELTKAMP:QUANTITATIVEMETHODSIN INDUSTRIALR&D
Whatever systems we apply and develop to measure R&D effectiveness and performance, one thing is sure, it is a continuous challenge and R&D management will constantly be exposed to one question: Are we doing the right things? - and are we doing the right things right?
462
Scientometrics 45 (1999)