Your AI-native startup ain’t the same as a typical SaaS company
Rudina Seseri, founder and managing partner at Glasswing Ventures, emphasized the unique challenges that AI startups face compared to traditional SaaS companies at the recent TechCrunch Early Stage event in Boston.
Seseri highlighted the misconception that merely integrating APIs qualifies a company as an AI company. True AI-native companies, according to Seseri, leverage algorithms and data as core components of their value creation.
She emphasized that customers and investors assess companies differently from SaaS startups. Unlike SaaS, where an unfinished product can be released and refined over time, products require a mature model before they can be trusted and adopted by customers.
Seseri pointed out the steep learning curve involved in training algorithms, highlighting the challenge for early-stage startups to strike a balance between model readiness and value creation for customers.
Finding early adopters for products is particularly challenging. Seseri advised startups to focus on articulating their value proposition to potential buyers, emphasizing problem-solving and business outcomes rather than technical intricacies.
She emphasized the importance of grounding discussions in business priorities and metrics, even while articulating a broader vision for the product. By aligning with customer needs and priorities, startups can navigate the complexities of early-stage adoption and build trust with their target market.
How can AI startups win?
Rudina Seseri emphasizes the importance of strategically positioning an AI startup within the competitive landscape, recognizing the challenges posed by established players who dominate various layers of the ecosystem.
In the analogy to the cloud era, Seseri identifies three layers: the foundation layer (housing large language models), the middle layer (where data processing and infrastructure are managed), and the application layer (where SaaS companies operate).
The foundation layer in AI is already heavily controlled by major players like OpenAI and Anthropic, making it difficult for new entrants to compete due to high capital requirements and the risk of commoditization.
However, opportunities exist at the application layer, where SaaS companies thrived despite the dominance of big players like Amazon, Google, and Microsoft. Seseri believes that startups can still carve out a niche in this layer, leveraging unique algorithms and data access to create defensible positions.
Additionally, there’s potential in the middle layer, where companies like Snowflake have succeeded by providing essential infrastructure for application players to manage their data.
Seseri advises investing in the application layer and selectively in the middle layer, prioritizing unique data access and algorithms as key differentiators. While building an AI startup presents significant challenges, particularly in comparison to SaaS startups, understanding the competitive landscape and strategically positioning the business can lead to success in this rapidly evolving industry.