Early-stage software investing being reshaped: Nexus Venture

As AI adoption accelerates, the benchmarks for early-stage software investing are being reshaped. Just a few years ago, $1 million in annual recurring revenue (ARR) was enough for a SaaS startup to go for a Series A round. Today, the new threshold, at least for top-quartile AI SaaS companies, is closer to $3-5 million, says Arjun Gandhi, vice president at Nexus Venture Partners. 

That said, the fund doesn’t use strict benchmarks, Gandhi adds. 

“If we love the team and their execution ability, we may invest pre-revenue with just a prototype. Other times, we’ll track companies for months to build conviction,” he says.

Rising benchmarks and global capital shifts

Adding to that, more Indian founders are raising funds early from US investors since there is more capital in that ecosystem, fetching higher valuations. They are also moving to the Silicon Valley soon after the initial funding round, to be closer to customers and the AI ecosystem. 

“While some of these valuations will look crazy in hindsight, others will look like bargains,” Gandhi added.

While the fundamentals of software investing remain the same, AI has changed the speed at which companies can scale, altering investor benchmarks. As development tools make product-building easier, distribution and go-to-market strategies have emerged as the most critical differentiators.

“At the application layer, the big differentiator is who understands their customer deeply and can sell effectively,” he said.

On this front, the firm is seeing new marketing playbooks gain traction, including viral campaigns that help startups cut through the noise. 

“As long as viral tactics aren’t offensive or rule-breaking, we view them positively. Everyone knows this is a once-in-a-decade paradigm shift, similar to the internet going mainstream,” Gandhi noted.

Enterprise focus, consumer AI potential

In India, nearly 70% of AI startups Nexus evaluates are at the application layer, largely targeting enterprise use cases. Model-layer companies, more common in the US, remain rare, though Gandhi sees long-term potential for more research-led ventures in India. In India, Nexus has invested in nearly 20 AI startups such as Neysa, Pratilipi, Cognida.ai, Jiffy.ai, and more.

Nexus, which is investing from its seventh fund of $700 million, primarily invests at the seed stage, with a few deals at a Series A and B.

While enterprise remains the dominant focus, Gandhi said consumer AI is an exciting frontier, with opportunities in AI companions, e-commerce, fashion and productivity tools, even though the firm is yet to take a bet on such propositions. 

Be it a consumer or an enterprise use case, AI is a “must-have”, not a “nice-to-have”, especially for Indian companies targeting global markets, primarily North America, Gandhi added. While newer companies have the advantage of being AI-native from day one, older companies often bolt on AI later since they can’t rebuild their architecture from scratch, which is less powerful.