Portfolio Management and Making the Right NPD and R&D Investments
1. A New AI Model Shows Promise in Predicting New Product Success
Dr. Robert G. Cooper
A pre-print article: Research-Technology Management, July-August 2025
Making the right investment decisions in new product development (NPD)—selecting the best projects to pursue—is essential to maximizing R&D productivity. Yet studies show that approximately 80% of these decisions prove to be wrong: many approved projects fail commercially or are later canceled, often after substantial investment.
To address this challenge, a new AI-driven model—AI-PRISM—has been developed by a leading researcher with decades of experience studying success factors in NPD. Under expert guidance, the AI was trained to build a decision-support model aimed at improving project selection outcomes.
The AI-PRISM model begins by analyzing the information submitted by the project team. It then supplements this with online research to fill in data gaps. Using this comprehensive information base, AI-PRISM evaluates the proposed project against 20 established criteria that distinguish likely winners from probable failures. It identifies the project’s strengths, weaknesses, and key risks, and suggests possible mitigation strategies. It also predicts the projects’ probability of success as a percentage, an estimate useful when doing financial analysis for the project.
AI-PRISM then generates detailed assessments to support the team’s Business Case. These include a market analysis, competitive landscape, Voice of Customer (VoC) summary, and technical assessment. It also recommends next steps and provides a data-supported Go/No-Go recommendation.
Beyond its ability to analyze vast amounts of information, AI-PRISM offers significant advantages: its evaluations are objective, consistent, and data-driven. For managers, this enhances decision-making professionalism and removes bias from critical Go/No-Go judgments. The model can also assess projects already in the portfolio, identifying candidates for potential discontinuation through an impartial “kill” recommendation.
For project teams, AI-PRISM delivers substantial value. It highlights project weaknesses and risks, recommends improvements, and provides insights for refining the Business Case—particularly in areas like market and technical assessments.
Citation:
Robert G. Cooper (2025). A New AI Model Shows Promise in Predicting New Product Success. Research-Technology Management, forthcoming July–August 2025: DOI: 10.1080/08956308.2025.2496109
PDF Download
2. AI-PRISM: A New Lens for Predicting New Product Succes
Dr. Robert G. Cooper, Feb 2025
Overview
For the first time ever, we can confidently recommend AI to predict new product success!
This is a game-changer. AI has finally cracked one of the toughest challenges in product development—identifying winners before costly investments are made.
Historically, up to 80% of approved NPD projects fail to become commercial successes (PDMA data). But now, our AI-PRISM model delivers a level of precision and reliability that human decision-making simply can’t match.
Introducing AI-PRISM—an autonomous AI-driven model that predicts new product success with remarkable accuracy. Developed through extensive testing and refinement, AI-PRISM leverages:
✅An AI-generated PRISM prediction model based on many success/failure studies
✅Automated information searching—market, technical, competitive info for the new product being analyzed
✅Validation through multiple independent runs to ensure maximum reliability.
All it takes is a simple project canvas (template) completed by the NP project team.
AI-PRISMdigests that canvass information, then searches vast data sources online to confirm and augment what the team provided. With all that information, the model rates the project on 20+ sub-factors across 7 key criteria. Using algorithms we developed with AI, the model delivers a success prediction and reasons why—faster and more reliably than managers or project teams alone.
Download the article, published in PDMA’s online journal in 2025.
To cite this article: Robert G. Cooper. 2025. “AI-PRISM: A New Lens for Predicting New Product Success.” PDMA Knowledge Hub (kHUB), Feb 28. Link: AI-PRISM: A New Lens for Predicting New Product Success
PDF Download
3. From Intuition to Insight: Leveraging AI to Forecast New Product Success
Dr. Robert G. Cooper, April 2025
Overview
Making the right RD&E investment decisions is critical for maximizing new product development (NPD) productivity. But up to 80% of such go/no-go decisions on NPD projects are incorrect. This article explores the underlying causes, highlighting the over-reliance on intuition rather than structured decision-making processes.
A new approach is the PRISM model created by Artificial Intelligence to tackle this tough decision. PRISM is designed to predict the likelihood of success of a new product project, based on seven empirically validated criteria/ Unlike traditional scorecards, PRISM calculates a probability of success, which when combined with financial data, offers management a more precise and value-based decision-making tool.
The article also examines how AI models like Perplexity Pro and DeepSeek can autonomously score projects using PRISM, with results showing strong reliability, more than twice that of human evaluators.
To cite this article: Robert G. Cooper. 2025. “From Intuition to Insight:
Leveraging AI to Forecast New Product Success,” IEEE Engineering Management Review, Early Access, April 14, pp. 1-10. DOI: 10.1109/EMR.2025.355977 Link: http://www.bobcooper.ca/images/files/articles/5/AIForecastsNPSuccessIEEEEMR.pdf
4. Strategies to Improve Portfolio Management of New Products
Given the substantial investments in R&D and the consequential impact on the business’s future, effective portfolio management decisions play a pivotal role in achieving business success. This article presents the findings of an Innovation Research Interchange (IRI) study on practices and outcomes in NP portfolio management. The various portfolio methods used, their popularity, portfolio breakdowns by project types, and portfolio performance results are all investigated, reported in easy-to-understand charts, and also compared to results from previous studies.
Key recommendations, based on the study and other similar reports tackle issues such as resourcing projects properly, getting the right mix and balance of projects in the portfolio, and improving portfolio performance. An insightful benchmarking report for any portfolio manager seeking to improve their firm’s portfolio management practices.
5. Expected Commercial Value for New-Product Project Valuation When High Uncertainty Exists
The most popular project valuation metric, the NPV, is wrong for capturing the true economic value for many new-product projects. By contrast, the Expected Commercial Value (ECV) provides a simple way to determine a project’s financial value where uncertainties and risks are high, as is the case for bolder innovations and Agile projects. New research provides a way to estimate the key likelihoods – commercial and technical success probabilities – needed in the ECV method. How this ECV method works in a stagewise new-product process model, and how to determine likelihoods of commercial and technical success are outlined with examples.
6. Dynamic Portfolio Management for New Product Development - Robert G. Cooper and Anita F. Sommer
Summary:Effective portfolio management is vital to maximizing the value of the business’ new product development (NPD) portfolio. But NP project evaluation and making R&D investment decisions is one of the weakest facets of NPD. One issue is that once a project is underway, much of the ongoing evaluation is backward looking and process focused: whether the project is on time and on budget, and reviewing checklists of deliverables. A second issue is that the project’s business case is rarely updated and reviewed in real time as the project progresses.
A dynamic portfolio management approach is one solution. The use of the Productivity Index (PI), which gauges the forward-value of an NPD project in real time, is key: the PI captures the project’s gain in value for every additional dollar spent. But financial metrics, such as the PI, often suffer from a lack of data integrity. Building in multiple iterations complete with demos to customers creates updated and more reliable data that greatly improve estimates of project financial value and the PI. Additionally, information with artificial intelligence (AI) and machine learning to undertake market analysis also enhances data integrity.
Since the value of a project depends on many factors, the use of a multi-dimensional Value-Based Scorecard is a useful tool -- a scorecard developed using proven, research-based criteria, and which captures more than just economic value. Properly used, the Value Based Scorecard greatly enhances the decision-making process at gates and at real-time “up-gates.” Examples are presented to illustrate key points.
Open Access article: The publisher, Taylor and Francis, has kindly made this an “Open Access’ article – free of charge to download:
https://www.tandfonline.com/doi/full/10.1080/08956308.2023.2183004
7. Value-Based Strategy-Reward-Win Portfolio Management for New Products
Traditional project valuation metrics, such as NPV, capture value to the shareholder. But today, with environmental, social, and governance goals so prominent, reducing the carbon footprint must be a factor in NP project selection. So should maximizing the customer’s experience. To meet the changing needs of companies, and based on its successful applications in some larger firms, the value-based scorecard (VBS) is introduced. The VBS is a multi-dimensional model based on three main value factors, S-R-W—strategic and mission, reward, and likelihood of winning—and ten items that make up these three value factors. Key charts to help decision makers better visualize their rating assessments are provided.
8. Unlocking “Pipeline Gridlock:” Effective Portfolio Management is the Key
Too many projects in the development pipeline is a common but serious complaint in new-product development departments. Pipeline gridlock leads to under-resourced development projects, which end up taking too long to get to market, and then often under-perform. Solutions are offered—Gates with Teeth, Red Flags, and the Productivity Index—to achieve a more balanced development pipeline with fewer projects but better projects.
9. New-Product Portfolio Management with Agile: Challenges and Solutions for Manufacturers Using Agile Development Methods
When Agile Development and the traditional way of measuring, evaluating, and managing projects clash, companies must reassess how their new-product portfolios are managed, how go/kill and prioritization decisions are made, and how project performance is measured. This article explores the new solutions and emerging challenges at the boundary between Agile project management and portfolio management, and suggests tools and approaches to deal with these new challenges.
10. Determining the Value of Ambiguous Agile Projects With Multiple Iterations Using Expected Commercial Value
Agile Development provides benefits to manufacturers, but also creates new challenges. One is how to place an economic value on Agile projects, which are ambiguous, uncertain, and fluid. The Expected Commercial Value (ECV) is outlined as a tool to gauge the economic value of highly ambiguous Agile projects with multiple iterations. The mathematical derivation of the ECV equations is described in detail.
11. We’ve Come a Long Way Baby and A Collection of Cooper’s Articles
This two-part article series begins with the introductory article “We’ve Come a Long Way Baby”, and is followed by “Collecting Cooper’s Articles”, which contains the other 17 articles Cooper has published in the Journal of Product Innovation Management.
When research into product innovation management began in earnest in the late 1970s, nobody knew much of anything about how to succeed in product innovation. Thus, the theme of much of the research in the years that followed was to probe the “drivers of success”. Some research looked at why new products win or fail; other studies lowered the microscope on businesses and their innovation performance, and sought reasons for their results, both positive or negative; and some focused on particular strategies and methodologies and their impact on performance. Read the complete file on this research and conclusions about what makes a winner in new-product development.
12. Your NPD Portfolio May Be Dangerous to Your Business’s Health
Cooper reveals data that points to dangerous trends in companies’ development portfolios – too many projects, an over-abundance of low-value projects, and too few significant development initiatives. He shows the reasons for this troubling trend, and then poses solutions, both strategic and tactical, to reverse the trend in your company.
13. Ten Ways to Make Better Portfolio and Project Selection Decisions
Cooper and co-author Edgett outline the ten best ways to make Go/Kill and project prioritization decisions in new-product development, and also ways to obtain the right mix and balance of projects in your development pipeline. A good “how to” article on a difficult topic.
14. Your NPD Portfolio May Be Dangerous to Your Business’s Health
Cooper reveals data that points to dangerous trends in companies’ development portfolios – too many projects, an over-abundance of low-value projects, and too few significant development initiatives. He shows the reasons for this troubling trend, and then poses solutions, both strategic and tactical, to reverse the trend in your company.
15. Effective Gates – Using Gates With Teeth
Product innovation productivity suffers from too many projects, with too few high-value ones, in most firms’ development pipelines. “Gates with teeth” help to prune the development portfolio of weak projects, and deal with a gridlocked pipeline. And a robust innovation strategy, coupled with strategic buckets, refocuses resources on high-value development initiatives.
16. Where Are All the Breakthrough New Products? Using Portfolio Management to Boost Innovation
This recent article in a special issue of Research-Technology Management outlines the best methods for making portfolio management decisions and R&D investment decisions in the case of riskier and bolder projects. The fact is that most project selection methods tend to favor small, incremental development projects, hence an overabundance of “renovation projects” in most firms’ development portfolios. See how to stop this trend to trivial new products, and rethink your investment decisions and methods. A good basis for anyone trying to redress the imbalance of projects – too many small, low value initiates in your development pipeline.