Amid global competition for economic dominance, innovation is central to success for companies, nations, and regions. It boosts productivity, improves performance, and underpins economic growth. AI enables companies to transform innovation productivity, but the breadth of potential use cases means companies often cannot see the complete picture and fail to maximize the benefits. In this Viewpoint, we explain how to radically improve innovation productivity by leveraging AI and a people-centric approach.
Productivity growth in developed economies is slowing, as the impact of aging populations and shrinking workforces begins to bite. Labor forces are getting older, hours worked per capita are falling, and the traditional engines of economic growth are losing steam. Manufacturing is especially hard hit (see Figure 1). We see evidence of this trend in Europe and Japan, but other advanced economies are realizing that continued prosperity depends on squeezing more output from every hour worked.
What is true for countries and regions is also valid at the company level. Manufacturers, engineering firms, process industries, and more are investing less in R&D. Simultaneously, innovation is becoming increasingly complex due to electrification, digitalization, engaging a wider ecosystem in innovation processes, and requirements of sustainability and cybersecurity regulations.
Companies are therefore spending more time on areas not directly aligned with their primary innovation objectives. In the not-too-distant future, this will lead to less differentiation and possibly to depressed profitability and further cuts to meaningful R&D. Avoiding this downward spiral is essential but daunting. Companies will need to do more with less, become more effective at achieving well-chosen innovation objectives, increase their innovation delivery speed, and avoid unnecessary time-consuming tasks. Put simply: innovation productivity must rise sharply.
At a high level, innovation productivity is output over input: realizing the biggest “bang for your buck” from innovation. More specifically, it is the ability to consistently generate high-impact innovation outcomes by improving effectiveness, quality, speed, and efficiency across the innovation lifecycle. This includes everything from identifying the right problems to executing solutions flawlessly and predictably (see the Innovation Productivity Framework in Figure 2).
Strengths and weaknesses within the Innovation Productivity Framework vary greatly between companies. However, common issues include speed of innovation (“disruptive competitors’ time to market is less than half of ours”) and efficiency (“our people spend half of their time on activities that have nothing to do with developing innovations for customers”). Dig deeper, and you find challenges around innovation effectiveness (“too often, we come up with innovations our customers are unwilling to pay for”) and quality (“we don’t spend enough time on innovation because half our time is spent fixing problems we should have avoided”).
There are many root causes for such problems, but they typically break down into organizational gaps, cultural barriers, and weaknesses in portfolio governance, processes, or team setup. In many cases, these factors interact — for example, poor portfolio and resource management can overwhelm teams, while unclear project ownership stifles execution. Similarly, weak project management processes lead to inefficiencies that may be misdiagnosed as resource shortages.
AI has the potential to transform innovation performance and drive higher productivity. The World Economic Forum (WEF) estimates that AI could increase economic growth by 4.4% over the next 10 years by accelerating innovation, with a further 2.5% increase in the subsequent decade — well above baseline forecasts.
Based on our recent research, culminating in the Arthur D. Little (ADL) report “Moving Innovation Forward,” we find that adopting AI can boost innovation productivity by 40% by selecting and implementing the right AI use cases, empowering people to innovate better and smarter. Examples (highlighted in the ADL Blue Shift report “Eureka: On Steroids!”) include:
Of course, AI is not a panacea, and it certainly is not one application. Our previous work found that there are close to 900 AI use cases across the innovation process, enabled by dozens of technical solutions produced by thousands of vendors.
This ocean of use cases makes it difficult for companies to navigate to those that will deliver the greatest impact, which can lead to the biggest missed opportunities. Often, the result is a piecemeal approach involving a large number of scattered use cases. Innovation departments can “fall in love” with an AI solution rather than taking the time to understand the real problem they need to fix. What’s needed is a practical, bottom-up approach combined with a strategic, top-down approach, all based on urgent needs and wider innovation objectives.
With more than 45% of the 900 AI use cases mapped in our study relevant to innovation productivity, making the right choices is crucial to delivering benefits at scale. Organizations must have a clear vision, a balanced AI portfolio, and a clear roadmap. This requires a focus on where to play, how to play, and how to scale.
Organizations need an innovation strategy that covers both incremental and breakthrough innovations, delivering initiatives that support both top-down strategic priorities and bottom-up AI use cases that drive productivity for the people involved in the innovation process.
This is the starting point for defining where to play (see Figure 3).
The strategic priorities should define the must-win battles to ensure that innovation productivity priorities align with the overall business and innovation strategy. This includes strategic choices across dimensions such as markets, product and service portfolios, competitive strategy, tech portfolio, and strategic execution.
For example, if growth and diversification are strategic priorities for a supplier of integrated logistics systems for warehouses, it may want to move to outcome-based business models (e.g., paid by number of parcels processed per hour). To do this, it would need to seamlessly manage the end-to-end handling process by collaborating with its customers and other onsite system providers. This has big implications for overall quality levels and speed/timing of its system engineering activities at the innovation productivity level, necessitating AI applications focused on system architecture, integration, modeling, simulation, and optimization.
To meet critical user needs, it is vital to understand the user personas within innovation teams.
Organizations must identify user personas such as chief engineering officer, mechanical engineer, and head of business development and their requirements then map these against available use-case clusters. Essentially, how can AI help create the “perfect innovator” for that persona? Use this approach to map and shortlist improvement opportunities, bearing in mind that AI is not the only method available to drive transformation. For example, AI can streamline complex project management, but it cannot replace skilled decision-making and the broad, deep human experience required to lead projects to success.
Figure 4 is an example of how the pain and gain points of a particular persona can be addressed, based on in-depth analysis from the “Moving Innovation Forward” report. In this case, a chief engineering officer needs to develop and maintain engineering standards and best practices, a pain point addressed by an AI use case that provides automated documentation. This people-centric approach and our use case database help individuals quickly navigate to the most powerful applications. By clustering applications sharing a common technical solution, individuals obtain maximum AI advantage at minimal solution complexity and costs.
The next task is identifying the most logical approach to transforming innovation productivity. How advanced does the technology need to be? Is it better to use a less-advanced, off-the-shelf large language model and train it on organizational requirements for use by a small group of innovators or to develop an in-house model from scratch for wider use?
The first step in this decision is creating a business case for investment, including how many people will benefit from the technology and what productivity gains might be broadly achievable. Business cases should include the cost of data access and the level of potential competitive advantage. For example, if a horizon-scanning tool that helps predict the compliance of a new product with national and regional regulations requires a multi-user subscription to an expensive third-party database, it may not be worth the investment.
In terms of scale, broader deployments are more challenging, but they may deliver higher benefits across the business. In addition, a widespread solution may not necessarily involve complicated and expensive digital solutions. Sometimes, simple off-the-shelf applications can create benefits if targeted at the highest-priority user needs.
Scaling the benefits of AI for innovation productivity is about more than technology; it is about people and ways of working. The Lab of the Future approach (see Figure 5), as set out in the Prism article “The People-Centric Lab of the Future,” puts people and their needs at its heart by focusing on:
Organizations should continually iterate between these four steps. For example, it is vital to involve IT early to ensure you can access the right data within corporate governance guidelines while ensuring regulatory compliance before looking to build data ecosystems. It is also important to pilot new applications to ensure they meet user needs. This process puts in place the right data, technology, people, governance, and processes for success. It enables accelerated scaling by leveraging the wider ecosystem to boost transformation, progress, and benefits.
We mentioned that organizations can realize a 40% improvement in innovation productivity by using a structured approach to implementing AI. Focused applications solve pain points for specific user personas, eliminating a scattergun approach that results in multiple pilot implementations that don’t scale or deliver value. By selecting use cases that maximize benefits against defined KPIs, success can be achieved with an comprehensive approach (see Figure 6):
This method — using AI to bridge the innovation productivity gaps that hold back many companies across regions and industries — delivers the benefits organizations need to compete in challenging markets, resulting in stronger companies and economies across the world.
Turbocharging innovation productivity is essential to competitiveness in a volatile world. AI offers the potential to transform innovation, but it must be applied strategically to meet well-understood user requirements. Our approach, which brings estimated innovation productivity gains of 40%, is based on answers to these questions:
By Michael Kolk, Philip van Basten Batenburg, Phil Webster , Martin Glaumann, Philipp Mudersbach, Shota Mitsuya
cloud technology
Este fondo, con un tamaño objetivo de 250 millones de euros, pretende...
