Optimizing R&D Investment and Making Strategic Decisions to Develop New Technological Opportunities
R&D Portfolio Analysis is a method for evaluating multiple research and development themes as a whole and determining how limited resources—including funding, personnel, facilities, and time—should be allocated. Rather than considering only the value of individual R&D themes, it examines relationships among projects, risk diversification, the balance between short-term and long-term objectives, and alignment with business strategy to maximize the overall value of R&D investments.
Corporate R&D involves a variety of activities, including improvements to existing products, the development of next-generation technologies, and the exploration of new technological principles. These activities have different objectives, time horizons, and levels of uncertainty, making it difficult to prioritize investments using a single set of evaluation criteria.
An important role of R&D Portfolio Analysis is therefore not simply to aggregate the evaluation results of individual research projects, but to determine which combination of R&D activities an organization should pursue.
R&D portfolio management must consider not only which projects deserve investment, but also which technological possibilities should be preserved and developed for the future.
Objectives of R&D Portfolio Management
The fundamental objectives of R&D portfolio management can be organized into three major categories: maximizing investment value, achieving portfolio balance, and maintaining alignment with corporate and business strategies.
Maximizing Investment Value involves allocating limited R&D resources to generate the greatest possible technological and business benefits. This requires considering not only the expected value of individual projects but also which combinations of projects can realistically be implemented.
Portfolio Balance concerns the appropriate combination of short-term and long-term research, improvements to existing technologies and exploration of new technologies, and projects with different levels of risk and expected returns. It also involves examining whether R&D investments are excessively concentrated in particular technological fields or business areas.
Strategic Alignment evaluates how R&D investments contribute to corporate, business, and technology strategies. This includes not only investments supporting current core businesses but also the development of technological capabilities needed to respond to future changes in business structures.
These objectives do not necessarily coincide.
For example, selecting only projects with high short-term returns may concentrate R&D resources on incremental improvements to existing businesses, leaving insufficient investment for future technologies and new business opportunities.
Conversely, allocating excessive resources to exploratory research may undermine product development and technological improvements needed to support current business operations.
R&D portfolio management must make these competing objectives explicit and determine an appropriate balance according to the organization’s circumstances.
Classification and Visualization of R&D Portfolios
R&D Portfolio Analysis commonly uses classification and visualization methods to examine multiple research themes according to selected evaluation dimensions.
One widely used approach is the portfolio matrix, which positions R&D themes along two evaluation axes. Examples include technical feasibility versus expected business value, technological novelty versus market uncertainty, and short-term profitability versus long-term growth potential.
Plotting research themes on such matrices helps identify areas where R&D investments are concentrated and where important capabilities or opportunities may be underrepresented. Bubble charts can also be used to represent investment amounts or resource requirements through the size of each plotted item.
However, two-dimensional matrices cannot always represent the multidimensional characteristics of R&D themes adequately.
For example, a project with low technical feasibility may still be worth exploring if an inexpensive experiment can resolve a critical uncertainty. Similarly, a technology with limited independent market value may be essential as a foundational capability for other research projects.
Portfolio analysis should therefore combine visualization with the evaluation of multiple criteria and relationships among research themes.
Major perspectives for examining R&D portfolios include the following:
| Analytical Perspective | Main Considerations |
|---|---|
| Technological Domains | Concentration in specific fields, cross-domain expansion, and diversity of technological capabilities |
| R&D Stages | Distribution across basic research, applied research, prototyping, and product development |
| Technological Novelty | Improvements to existing technologies, new applications, and exploration of breakthrough technologies |
| Time Horizons | Short-term outcomes, medium-term development, and long-term research investments |
| Business Domains | Existing businesses, new product areas, and expansion into new markets |
| Uncertainty | Distribution of technological and market uncertainties and associated risks |
| Resource Allocation | Allocation of budgets, personnel, facilities, and research time |
| Project Relationships | Shared technologies, complementarities, dependencies, overlaps, and competition among projects |
Combining these perspectives enables organizations to examine the overall structure of R&D investments, which cannot be fully understood through simple project rankings.
Exploitation of Existing Technologies and Exploration of New Possibilities
Corporate R&D requires both activities that maintain and strengthen the competitiveness of existing businesses and activities that create opportunities for future growth.
The exploitation of existing technologies focuses on relatively well-defined objectives, such as improving product performance, enhancing quality, reducing manufacturing costs, and increasing technological reliability. These activities are essential to sustaining current business operations and revenues.
In contrast, the exploration of new technological possibilities investigates unfamiliar technological principles and knowledge from different disciplines to identify potential future products and business opportunities.
Exploratory R&D generally involves greater uncertainty, making it difficult to predict financial returns in the short term.
Consequently, when projects are compared primarily through current profitability or expected payback periods, exploratory research may receive lower priorities than incremental improvements.
One approach to addressing this issue is the Strategic Buckets method.
Under this approach, organizations establish resource allocation categories corresponding to different strategic objectives, such as improvements to existing products, next-generation technology development, and long-term exploratory research.
R&D themes are then evaluated and selected within the relevant categories.
This helps avoid forcing fundamentally different research activities into a single ranking system and can protect investments in long-term technological capabilities.
However, investment proportions should not necessarily remain fixed. Changes in technological developments, competitive environments, and business conditions may require revisions to the composition of the portfolio.
Moreover, knowledge generated through exploratory research may contribute to improvements in existing businesses, while established technological capabilities may enable expansion into unfamiliar fields.
Portfolio management should therefore recognize relationships among different categories of R&D activities rather than treating them as entirely separate.
Interdependencies and Synergies Among R&D Themes
One of the most important issues in R&D Portfolio Analysis is that individual research themes are not necessarily independent.
Consider the development of a new product that requires advances in materials, manufacturing processes, and measurement and control technologies.
Each of these areas involves distinct technical challenges. However, even if a new material is successfully developed, commercialization may remain impossible without appropriate manufacturing processes and quality assurance technologies.
Conversely, a measurement technology developed for one project may be applicable to several other research themes, increasing the overall value of the R&D portfolio.
Relationships among R&D themes can be classified into several categories.
Complementary Relationships occur when multiple technologies create functions or value that cannot be achieved independently.
Dependency Relationships exist when the feasibility of one project depends on the technological outcomes or capabilities developed through another.
Shared Technology Platform Relationships arise when a technological capability or research outcome can support multiple R&D themes.
Alternative or Competitive Relationships occur when projects propose different solutions to the same technical problem or compete for limited research resources.
Knowledge Spillover Relationships occur when knowledge acquired through one research project contributes to the development of other projects.
When such relationships exist, simply adding the evaluation scores of individual projects may fail to represent the actual value of the portfolio.
Interactions among research themes are not always positive. Competition for resources, technological duplication, and dependence on shared technical assumptions can increase costs or risks across the portfolio.
Therefore, R&D Portfolio Analysis must evaluate not only the value of individual projects but also the additional value and constraints arising from their combinations.
Multiple-Criteria Decision Analysis and Portfolio Optimization
R&D themes are evaluated using multiple criteria, including technical feasibility, expected business value, strategic alignment, technological novelty, and uncertainty.
Multiple-Criteria Decision Analysis (MCDA) provides methods for comparing alternatives across these different dimensions.
Methods such as AHP, weighted scoring, TOPSIS, ELECTRE, and PROMETHEE enable organizations to evaluate research themes while considering multiple criteria.
However, evaluating individual R&D themes using multiple criteria is different from optimizing the overall portfolio.
For example, selecting projects in descending order of their evaluation scores may lead to excessive investment in a particular technological field. It may also produce an infeasible portfolio because of limitations in research personnel, equipment, or other resources.
Portfolio optimization treats project selection as a problem of choosing combinations of research themes under constraints involving budgets, personnel, facilities, development schedules, and other requirements.
Mathematical programming can be used to identify combinations of projects that maximize an evaluation function under specified constraints.
Multi-objective optimization can also consider competing goals such as expected value, technological diversity, risk, and strategic alignment.
However, a mathematically optimal solution is optimal only with respect to the evaluation model and constraints that have been specified.
If the underlying evaluations or future assumptions are uncertain, the resulting portfolio should not be treated as an unquestionably correct investment decision.
Sensitivity analysis and scenario analysis are important for examining how portfolio selections change under different assumptions and future conditions.
Evaluating Interactions with the Choquet Integral
Conventional weighted scoring assigns a weight to each evaluation criterion and calculates an overall score through a weighted average.
This approach is relatively easy to understand and implement, but it may not adequately represent interactions among evaluation criteria.
For example, high technical feasibility and strong market potential may create additional value when both are present. Conversely, several evaluation criteria may represent overlapping characteristics, leading to double counting when their scores are simply added together.
The Choquet Integral is an aggregation method that can account for such interactions by assigning non-additive weights to combinations of evaluation criteria.
The underlying concept can also be extended beyond the evaluation of individual projects to the assessment of entire R&D portfolios.
Brunelli and Corrente (2024) proposed a multiple-criteria portfolio decision model that considers both interactions among evaluation criteria and interactions among projects.
Their approach uses the 2-additive Choquet Integral, which primarily represents pairwise interactions, to balance modeling flexibility with computational and practical manageability.
Such an approach may be useful when combinations of R&D themes create additional value through technological synergies or when duplication and competition reduce the overall value of a portfolio.
However, applying the Choquet Integral in practice presents several challenges.
The interactions to be modeled must be identified, and their parameters must be estimated or elicited from decision-makers. As the number of projects and evaluation criteria increases, the amount of information required may also grow substantially.
Furthermore, technological dependencies, project sequences, and resource constraints may require integration with other analytical methods or optimization models.
The objective is not to employ sophisticated mathematical techniques for their own sake, but to understand meaningful relationships among R&D themes and incorporate them appropriately into strategic decisions.
Dynamic Portfolio Management Under Uncertainty
An R&D portfolio cannot remain fixed indefinitely.
As research progresses, new information becomes available concerning technical feasibility, development time, required resources, and potential market value.
This is particularly important in exploratory R&D, where initial hypotheses may be rejected or unexpected technological discoveries may emerge.
Portfolio management must therefore involve continuous reassessment.
For example, an organization may initially allocate limited resources to an exploratory project to test a critical technical assumption.
If the results are promising, additional investment may be justified. If the assumption cannot be validated, the project may be redirected or terminated.
Such staged decision-making can be supported by Stage-Gate management and real options reasoning.
However, portfolio reassessment should not be limited to decisions about whether individual projects should continue.
The effects of new research findings on other projects must also be considered.
Even if a project fails to achieve its original objective, the knowledge acquired may prove valuable in another technological domain.
Similarly, discovering a new technological possibility may change the strategic importance of existing research themes.
R&D portfolio management should therefore monitor not only project progress but also the accumulation of technological knowledge and the emergence of new research opportunities.
Connecting Concept Exploration with R&D Portfolio Analysis
Conventional R&D Portfolio Analysis generally assumes that a set of candidate research projects has already been identified.
Its primary tasks are to evaluate, select, and allocate resources among these candidates.
However, when new research themes emerge through the acquisition and recombination of technological knowledge, the set of projects available for evaluation is not fixed.
For example, cross-domain exploration of a technological problem may reveal several previously unrecognized solution principles.
Some of these principles may represent alternatives to existing research projects, while others may complement existing themes or create entirely new research opportunities.
Furthermore, combining several hypotheses may lead to the conception of technological systems that could not have been anticipated from any single hypothesis.
In such cases, technological hypotheses must be organized and their relationships analyzed before portfolio evaluation.
At least three distinct levels of analysis should be considered:
| Level of Analysis | Main Objects | Analytical Purpose |
|---|---|---|
| Technological Elements | Functions, structures, operating principles, and technological constraints | Understanding technological knowledge and supporting cross-domain exploration |
| R&D Hypotheses | Proposed solution principles, applications, and conditions for feasibility | Integrating and comparing hypotheses and identifying verification requirements |
| R&D Themes | Research objectives, implementation plans, required resources, and expected outcomes | Investment evaluation, prioritization, and portfolio construction |
A single technological element may contribute to several hypotheses, while multiple hypotheses may be integrated into one R&D theme.
Conversely, a particular technological hypothesis may provide a common foundation for several different R&D themes.
Analyzing these multilayered relationships may reveal shared technological capabilities, complementarities, dependencies, and opportunities for coordinated research.
However, technological similarity alone does not demonstrate that projects have practical synergies or dependencies.
The conditions under which such relationships hold must be examined through technical evidence and expert assessment.
Connecting Concept Exploration with R&D Portfolio Analysis offers a way to extend portfolio management beyond the selection of existing projects toward the continuous creation and evaluation of new R&D opportunities.
The R&D Portfolio Analysis Process
R&D Portfolio Analysis can be organized into the following process:
| Stage | Main Activities |
|---|---|
| 1. Strategic Objective Definition | Clarify business strategies, technology strategies, and short-term and long-term R&D objectives |
| 2. Theme Identification and Classification | Organize ongoing and planned R&D themes according to objectives, technological domains, maturity, and other characteristics |
| 3. Individual Theme Evaluation | Assess technological value, business value, strategic alignment, uncertainty, and resource requirements |
| 4. Analysis of Project Relationships | Examine shared technologies, complementarities, dependencies, overlaps, and competition |
| 5. Portfolio Construction | Evaluate combinations of R&D investments based on strategic objectives, resource constraints, and project relationships |
| 6. Scenario and Sensitivity Analysis | Examine the robustness of portfolio choices under alternative assumptions and future conditions |
| 7. Investment Decisions and Continuous Review | Allocate resources and revise the portfolio in response to research outcomes and environmental changes |
In practice, these stages should be repeated as research progresses and new technological opportunities emerge.
R&D Portfolio Analysis should also be connected to technology roadmapping and research planning.
It is not sufficient to decide which technological domains deserve investment. Organizations must also clarify which technologies need to be established, when they will be required, and which research outcomes serve as prerequisites for subsequent development activities.
Mindware Research Institute’s Approach
At Mindware Research Institute, we regard R&D Portfolio Analysis not merely as a problem of optimizing investments among existing research projects, but as a process of expanding the set of available R&D opportunities through technological exploration and developing R&D strategies based on the relationships among those opportunities.
Conventional portfolio analysis often assumes that the candidate R&D themes are already known.
However, the research themes currently recognized by an organization do not necessarily represent all the technological possibilities worth investigating.
ConceptMiner R&D, a Cross-Domain R&D Concept Exploration System developed by Mindware Research Institute, is designed to extract knowledge from technical literature, abstract technological characteristics and structures, explore different technological domains, and generate new research hypotheses.
ConceptMiner R&D combines LLM-based technological knowledge analysis with concept mapping using machine learning techniques such as Growing Neural Gas (GNG) to help researchers understand the conceptual structure of technical information and generated hypotheses.
A potential direction for further development is to analyze the relationships between generated hypotheses and the technological elements from which they were derived, extending Concept Exploration toward R&D theme formation and the analysis of relationships among research themes.
For example, identifying technological elements shared by multiple hypotheses may reveal candidates for foundational technologies that could support several R&D themes.
Similarly, examining the conditions required for a hypothesis to hold may help identify dependencies among research projects and determine which technical challenges should be verified first.
However, conceptual proximity on a map does not necessarily imply technological synergy.
Similarity analysis must be distinguished from the analysis of actual complementarities, dependencies, and other causal or functional relationships.
By connecting these analytical capabilities with MCDA and portfolio optimization, it may become possible to support a more integrated process extending from R&D theme creation to evaluation, prioritization, and resource allocation.
Under this approach, an R&D portfolio is not simply a fixed collection of previously approved projects.
Instead, it is treated as a dynamic set of research opportunities that evolves through knowledge acquisition, hypothesis generation, technological verification, and changes in the business environment.
Final portfolio composition and investment decisions must, of course, reflect the organization’s specific strategies, resources, technological capabilities, and market conditions.
Human judgment remains essential.
The role of AI and mathematical analysis is not to replace executives or R&D managers in making investment decisions, but to reveal relationships among projects, examine alternative combinations of investments, and provide information that supports more informed decisions.
The essential purpose of R&D Portfolio Analysis is not merely to select the optimal combination from currently available research themes. It is to continuously discover new technological possibilities and develop future R&D options by connecting them with the organization’s capabilities and business strategies.
Mindware Research Institute aims to connect the creation of new R&D themes through Concept Exploration with portfolio analysis that considers relationships among research themes, supporting a more integrated approach to knowledge creation and strategic decision-making in research and development.
Related Topics
What Is R&D Management?
An overview of R&D strategy, theme exploration and evaluation, resource allocation, and commercialization.
Technology Intelligence
Collecting and analyzing technological information to identify new research and development opportunities.
R&D Theme Exploration and Creation
Generating R&D hypotheses through Concept Exploration, structural analogy, and cross-domain knowledge exploration.
R&D Theme Evaluation and Prioritization
Evaluating, integrating, and selecting individual R&D hypotheses and themes and developing verification plans.
Technology Roadmapping
Connecting technological developments, R&D investments, product strategies, and commercialization plans over time.
Multiple-Criteria Decision Analysis (MCDA) and AHP
Methods for comparing and evaluating R&D themes against multiple criteria.
Choquet Integral
A method for aggregating evaluations while accounting for interactions among criteria, with potential applications to non-additive portfolio evaluation.
References and Further Reading
- Cooper, R. G., Edgett, S. J., & Kleinschmidt, E. J. (1997). Portfolio Management in New Product Development: Lessons from the Leaders—I. Research-Technology Management, 40(5), 16–28.
- Cooper, R. G., Edgett, S. J., & Kleinschmidt, E. J. (1997). Portfolio Management in New Product Development: Lessons from the Leaders—II. Research-Technology Management, 40(6), 43–52.
- Cooper, R. G., Edgett, S. J., & Kleinschmidt, E. J. (1998). Best Practices for Managing R&D Portfolios. Research-Technology Management, 41(4), 20–33.
- Brunelli, M., & Corrente, S. (2024). Modeling Criteria and Project Interactions in Portfolio Decision Analysis with the Choquet Integral. Omega, 126, 103076.
- Vieira, G. B., Oliveira, H. S., Almeida, J. A., & Belderrain, M. C. N. (2024). Project Portfolio Selection Considering Interdependencies: A Review of Terminology and Approaches. Project Leadership and Society, 5, 100115.
- Solak, S., Clarke, J.-P. B., Johnson, E. L., & Barnes, E. R. (2010). Optimization of R&D Project Portfolios Under Endogenous Uncertainty. European Journal of Operational Research, 207(1), 420–433.