Currently, an estimated 70% of new product developments approved for development never become a commercial success, Dr Robert (Bob) G. Cooper writes in his article ‘The AI transformation of product innovation’.
He observes that these projects either fail in the marketplace or are killed before launch, often after considerable financial investment; he admits that “tossing a coin would give better decisions!”
He sees a key role for AI in developing a robust business case to support good go/no go decisions, potentially making significant savings for companies (and their investors) and accelerating successful innovation .
How to win at new product development
There are two ways to win at new products:
- Do the projects right – employ voice-of-customer research, use best practices, and assemble an experienced cross-functional team.
- Do the right projects – create a strong business case, ensure strategic alignment, and make rational decisions based on key success criteria.
The ‘go-to-development’ decision is one of the most resource-intensive commitments in new product development (NPD) but also the most error-prone.
Bob explains that an important concept in NPD is that it is a process, and like any process, it can be managed to make it faster, more effective, and more productive. He was the originator of the Stage-Gate process, which identifies a series of stages in NPD, each with defined tasks designed to gather information that will reduce risk and uncertainty, and decision nodes or gates for monitoring progress and making the Go/No-Go decisions to move to the next stage.
“If one views NPD as an information process then AI is ideally poised to power this process,” he explains, and takes the example of Voice of the Customer (VoC).
AI for VoC makes NPD more productive
To understand customer needs and pain points, firms often conduct in-depth qualitative interviews, but these are expensive and difficult to schedule. Also, after a few interviews, the same issues will surface.
AI tools for VoC can be used to gather and analyse online text from user forums, reviews, complaints, and magazines to reveal what customers like and dislike and to identify unmet needs.
This information could be used to optimise the value of the in-person interviews or to see patterns and behavioural insights that a human may miss.
This rich information, when combined with an analysis of competitors’ activities, market trends, and sales volumes, would provide a much more detailed business case for NPD than historically available.
Furthermore, it could be used to create 3D renderings of product concepts that make it easier for stakeholders to see what might be involved in developing the product and to gain customer reactions in a concept test.
The article describes how General Motors works with MIT to employ a generative AI model that can forecast consumer preferences for car design, removing the need for ‘theme clinics’ that cost $100,000 and involve hundreds of consumers.
Better go-to-development decisions
The ‘go-to-development’ decision is a natural candidate for AI. AI can:
- Analyse vast internal datasets, spotting patterns to do market and sales forecasts
- Undertake online market, competitive, and technical searches, accessing far more sources than a project team would, and summarizing the results into a readable document
- Write a robust business case by working interactively with the project team
- Help to eliminate management bias.
The management team, informed with better information, is then poised to make better go/no-go decisions.
AI can do much more than provide better information for the go-to-development decision. In another article, Bob outlines the AI-PRISM success prediction model. Here the AI model reviews the project’s business case prepared by the project team, and then fills the information gaps by doing a thorough online market, competitive, and technical search.
Next, AI-PRISM, using both sets of data, autonomously scores the project on 20 key success criteria, and determines the probability of commercial success for the new product. It also does a strengths/weakness assessment of the project, and comments on the projects’ prospects. The model has been validated on many real projects and yields more reliable results than management teams when faced with these tough “go to development decisions”.
Another example is Sopheon’s Accolade, which optimises a firm’s development portfolio, prioritising projects, given the resources and skills available. Sopheon’s Optimizer model simulates different portfolio options, predicting results in terms of strategic alignment, profitability, and resource requirements.
A Forrester study estimated that through optimised project management, there was potential to accelerate time to market by 15% and save 10% of the project budget through improved resource utilisation.
Building a strong business case for AI
Some firms are reluctant to adopt AI due to the lack of a strong or ‘provable’ business case. Bob agrees there are few rigorous large-scale studies of companies using AI but says that there are smaller but robust studies that strongly support the business case for new product development.
“The benefits of a new technology are always difficult to predict, but early adopters cite improved efficiency, business agility, and productivity.”
“However, the number one benefit realised is increased innovation, with significant pay-offs. Some quote achieving a 50% reduction in development time.”
“As the adoption window for new technology becomes progressively shorter, now is the time to act.”
“Get up to speed on AI in NPD, get some outside help, put a plan in place, and undertake some pilots. The only way to escape obsolescence is to embrace innovation.”
AI for Innovation – 7th May 2025
Robert Cooper will be speaking at the AI for Innovation virtual conference on 7th May 2025.
Aimed at those tasked with innovation strategy and new product development across all industry sectors, this pragmatic conference aims to provide context for recent AI developments, offer insights from data scientists, and share real-world industry experience from those who have started their journey.
The interactive format will enable you to explore the potential of AI for your organisation, talk to practitioners and develop an action plan.
More information and to register.
Read the papers:
‘The AI transformation of product innovation’, Industrial Marketing Management, Volume 119, 2024
‘AI-PRISM: A New Lens for Predicting New Product Success’, PDMA Knowledge Hub (kHUB), Feb 28 2025
