Stellaris Venture Partners reposted this
Had a great time at the GSF Founders-Investors Conference! Thanks Rajesh Sawhney for this opportunity & Khadim Batti for your insights! Two key topics of discussion, both relevant for Indian #SaaS entrepreneurs: 1. Domiciling your company 2. Approach to AI Few thoughts: ➡ Domiciling your SaaS company in the US has been the norm. Why? - US customers want that - US investors want that - Future acquirers want that - US IPO is the preferred exit aka Freshworks Most of these assumptions are not 100% true. US customers are OK as long as contracts are being written from a US entity. That entity can be a subsidiary of your Indian parent. I believe this is not a practical challenge. Yes, US investors wanted this, but it has changed substantially. When Whatfix was raising its Series B, I remember that a few US investors "preferred" a US HQ. Today, most US investors are OK with an India entity. Y Combinator also recently has started funding companies HQ'ed in India! For mid-sized acquirers, having IP in the US is still a plus, but for larger acquirers (e.g. SAP, Oracle, Cisco), this is not an issue. Their M&A machines are well oiled to make acquisitions across the world. This is the most significant reason as to why you want your company to be HQ'ed in India now. US IPO is quite hard, the bar is $250-300M at least, and even then, you need to truly stand out in your category. India, however, may offer a scarcity premium for SaaS companies to be listed. We have seen RateGain go public in India. I expect several other IPOs in India. One can go public in India with even $50M ARR, and still be a highly differentiated offering! ➡ On AI, loved Khadim's quote: In the short term, we are all overestimating AI's impact & in the long term, we are all underestimating it! Incremental features like natural language based analytics, autocomplete, summarization are important, but every provider needs to ask themselves if they can reimagine the category with AI. At Whatfix, the team is rethinking what human machine interface should be. Instead of humans adapting to applications, it is applications that will adapt to humans, which means that applications in the future may be "invisible". Humans can instruct in natural language what they want to get done from a place they want to e.g. Slack, and expect the software to autoexecute through agents. ➡ That being said, we are not there today. As Khadim said, we are overestimating what AI can do today. AI has enough hallucinations, and for enterprises, "almost OK" is not "OK". The level of accuracy required is far off. In the short term, we will see higher investments in product development as software providers raise the bar with reimagination of their categories. In the longer term, we expect costs of development to come down as AI makes it easier, faster and cheaper to build! If you're a founder reimagining an existing category or creating a new one, we would love to speak with you! Stellaris Venture Partners