Artificial intelligence startups are unlikely to replicate the high-margin economics that defined the software-as-a-service (SaaS) boom, according to venture capital firm Kalaari Capital, which is urging founders to recalibrate their strategies in response to a fundamentally different cost structure.
In an article titled “AI startups don’t enjoy SaaS-era margins — here’s what Kalaari Capital’s Harshit Kumar says founders should do instead,” published by The Economic Times, Kalaari Capital’s vice president Harshit Kumar argues that the economics underpinning AI businesses diverge sharply from traditional SaaS models. The shift, he suggests, demands a rethinking of how founders build, price, and scale their products.
Unlike SaaS companies, which benefited from relatively predictable and declining marginal costs as they scaled, AI startups face persistent expenses tied to compute, model training, and inference. These ongoing costs limit gross margins and complicate the path to profitability. Kumar emphasizes that founders should avoid assuming that AI businesses will naturally achieve the same financial efficiency that made SaaS such an attractive category for investors over the past decade.
Instead, he advocates for a more disciplined approach to building AI companies. This includes focusing on high-value use cases where customers are willing to pay a premium, rather than pursuing growth through commoditized offerings. Founders are also encouraged to tightly align pricing with the value delivered, ensuring that revenue scales in tandem with operational costs.
Another key recommendation involves prioritizing strong distribution and customer integration. AI products that are deeply embedded into customer workflows are more likely to command pricing power and sustain usage, helping offset the higher cost base. Kumar also points to the importance of optimizing model usage and infrastructure efficiency to manage expenses more effectively.
The article highlights a broader shift in investor expectations. While SaaS companies were often evaluated on growth metrics and recurring revenue, AI startups are increasingly judged on their ability to demonstrate clear monetization pathways and cost discipline. This reflects a more cautious funding environment, where capital is less forgiving of business models that rely on margin expansion assumptions that may not materialize.
Kumar’s perspective underscores a growing consensus within the venture community: AI represents a transformative technological opportunity, but one that does not conform to the economic playbook of earlier software cycles. Founders who acknowledge these constraints and adapt their strategies accordingly are more likely to build sustainable businesses in a competitive and rapidly evolving market.
