More than technology: what investors look for in AI startups
The AI Pitch Competition offers startups a platform. What do investors really look for in early-stage investments? Two experts open up.
Published on August 14, 2025

As Head of Partnerships, Linda liaises with new partners. She coordinates all ongoing collaborations and connects our journalistic editorial and commercial articles. She is not only the connector behind our articles, but also all of our events.
Artificial intelligence has become an integral part of our society and economy. The technology has the potential to disrupt almost every sector, offering plenty of opportunities for startups. But how can an AI startup become a successful company? What do investors look for in the early stages? Andy Lurling, founder of LUMO Labs, and Lotte Smit van Ditshuizen, partner at Arches Capital, share their insights on the opportunities and pitfalls for AI startups.
Impactful and unique
“It may sound strange to an investor who focuses on deeptech, but technology alone is not enough,” says Lurling. “It's about the problem you're solving. Technology is the means, not the end.” For LUMO Labs, the positive impact of the startup is crucial when making new investments. Lurling: “It's about the long-term impact. We always invest in trends, not hype. Hype can help a startup grow quickly, but it's often not sustainable in the long term. A trend, on the other hand, can bring about change.”
Smit van Ditshuizen agrees: “The product must be truly unique. Developments in AI are happening at lightning speed, making products relatively easy to copy. Having access to unique data is essential for success, especially when facing the data advantage of big players.” She mentions Flatiron Health as a successful example. “This AI startup in oncology had exclusive access to millions of anonymized patient records. That unique dataset made the company indispensable for pharmaceutical companies and researchers, and ultimately led to its acquisition by Roche for $1.9 billion.”
With a unique and distinctive product, the chances of lasting success are greater, she asserts. “If you build something that solves a major customer problem and that users can't live without once they've tried it, you've struck gold.”
Good customer relations
The speed of developments means that startups need to secure customers at an early stage, Lurling observes. “You don't want customers to be able to switch to another provider on a weekly or monthly basis. It can help to build a business model based on licenses. Above all, it's important to build a good relationship with the customer. If a startup is a reliable partner, the threshold for switching becomes higher.”
Niche markets
Most opportunities for AI startups lie in niche markets. Smit van Ditshuizen: "Large companies invest enormous amounts in AI. If you, as a startup, follow their roadmap too closely, it becomes almost impossible to differentiate yourself and stay ahead. Therefore, focus on a niche and build in-depth expertise in a specific domain, which will enable you to make a difference. A good example is Tractable, which focuses entirely on AI for damage inspection of cars and property. They are the global market leader in this field. To compete with large tech companies, access to unique data is also essential, as is the case with Flatiron Health."
Keep talking
Lurling and Smit van Ditshuizen see talking to partners, potential customers, and others in the target market as the most important aspect of running a successful startup.
“Don't be afraid to start talking to different stakeholders with just a prototype. You learn from it and it ultimately makes your product better,” says Lurling. Smit van Ditshuizen adds: “That way, you learn where the real pain points are for potential customers and discover the ideal customer profile for your product.”
One thing is certain: building an AI startup does not happen overnight. Entrepreneurs face a wide range of challenges. “The biggest challenge lies in the quality of the data and the rules surrounding data sharing,” says Lurling. “When you work with AI, the data that the algorithm works with is crucial. If the quality of the data is suboptimal, so is the end product.” In addition, it is also important for AI startups to look at the rules for sharing data. “It is crucial to consider this carefully so that you don't encounter any surprises along the way.”
AI Pitch Competition
Lurling and Smit van Ditshuizen see the AI Pitch Competition as an important way for startups to increase their visibility. The AI Pitch Competition is for early-stage AI startups in the early stages. Participants must have identified a customer segment and found the best solution for the problem in that customer segment. The AI Pitch Competition can then help you refine your pitch and tap into a new network to take steps toward follow-up financing. Lurling: "People must know your company. That ultimately increases reliability. Would you like to participate in the AI Pitch Competition with your startup? You can register until August 29 via this link.
Sponsored
This story is the result of a collaboration between AI Pitch Competition and our editorial team. IO+ is an independent journalism platform that carefully chooses its partners and only cooperates with companies and institutions that share our mission: spreading the story of innovation. This way we can offer our readers valuable stories that are created according to journalistic guidelines.
Want to know more about how IO+ works with other companies? Click here