AI-powered ideation platform at Hempel

Hempel had a well-established innovation setup where they systematically developed and matured ideas through prototyping and conceptualization. The biggest bottleneck in the process was in the early stages: generating enough ideas, and not least, good ideas, to fill the so-called innovation pipeline. The manual processes of idea workshops and interviews with internal and external stakeholders were both resource-intensive and limited.

As a result, the potential of the rest of the innovation process was rarely fully realized. Hempel needed a solution that could increase the quantity and quality of idea input - without compromising on business understanding and user needs. At the same time, Hempel sought to integrate AI as a strategic lever in its innovation work, without losing sight of its human and market-oriented starting point.

“Coming up with a few good ideas is rarely the problem. The real challenge is generating a high volume of relevant, high-quality ideas consistently, especially when you're trying to solve complex problems or break new ground." - Nicolai Hansen, Senior Innovation Manager, Hempel.

In close collaboration with Delegate, Hempel has developed IdeaLab - a digital platform that combines generative AI with synthetic personas and business logic. Built on Microsoft Azure, the solution offers a user-friendly, web-based interface where innovation professionals can define a challenge, select relevant market segments, and activate virtual personas to generate ideas.

IdeaLab is built on an advanced integration of generative AI and business-oriented design methodology, where each digital persona is modeled based on specific customer profiles and behavioral data. This makes it possible to simulate realistic and nuanced reactions to ideas - as if they were assessed by real users in different segments. The underlying AI uses natural language processing and semantic understanding to generate and evaluate ideas, ensuring high relevance and contextualization. The system is also scalable and modular, allowing new personas and business scenarios to be easily added, enabling Hempel to continuously adapt and evolve the platform in line with changing strategic needs. The combination of automation and human-like feedback significantly reduces both time and complexity in the early stages of innovation.

The AI then generates a series of ideas that are automatically evaluated by synthetic personas—digitally simulated users who provide feedback and score the ideas. The top-rated ideas undergo an AI-supported process where the concept is refined using tools such as the Value Proposition Canvas and Business Model Canvas. Outputs are generated in presentation format and can be directly used in management decisions. Hempel is already using the platform in its innovation workshops and has plans to roll it out to wider business units.

With IdeaLab, Hempel has transformed its approach to innovation. Where ideation previously required lengthy, manual processes, the company can now generate and qualify ideas at far greater scale and speed. The result is an 8-fold increase in the speed of idea generation, 320% more ideas, and a 270% increase in ideas ready for the next stage of the innovation process. In addition to quantitative improvements, the innovation department finds that ideas are of higher quality and more relevant to both the market and end users. Additionally, synthetic personas offer an extra layer of validation, enabling quick and easy assessment of the impact of ideas across target groups. IdeaLab has not only streamlined the innovation process, it has also strengthened Hempel's ability to innovate, experiment and develop future solutions in a competitive global market.

The significant impact of IdeaLab is due to a systematic combination of automated idea generation and intelligent idea maturation, driven by machine learning algorithms and business rules. Instead of starting from a blank canvas, the platform leverages historical data, market trends and industry insights to suggest ideas with a high probability of business relevance. As ideas are generated, multiple parallel evaluation streams are activated via digital personas programmed to represent different customer types, markets and needs. This creates a simulated market response early in the process, reducing reliance on extensive manual user studies. The AI-supported maturation phase enables ideas to be refined using established innovation frameworks, such as Design Thinking and Lean Startup, ensuring that the output is not only innovative but also grounded in realistic business models.