Most SaaS startups begin with SMBs for faster sales cycles. Nexla did the opposite, targeting complex enterprise problems from day one. This forced them to build a deeply capable platform that could later be simplified for smaller customers, rather than trying to scale up an SMB solution.

Unlike the slow denial of SaaS by client-server companies, today's SaaS leaders (e.g., HubSpot, Notion) are rapidly integrating AI. They have an advantage due to vast proprietary data and existing distribution channels, making it harder for new AI-native startups to displace them. The old playbook of a slow incumbent may no longer apply.

While platform businesses (marketplaces) can achieve massive valuations, they are incredibly difficult and expensive to build due to the chicken-and-egg problem. For most founders, a traditional B2B SaaS model is a far safer and more direct path to success.

In the previous SaaS era, emulating giants like Salesforce was a common but flawed strategy for startups. In the new AI era, there is no playbook at all, forcing founders to rethink go-to-market strategies from first principles rather than copying incumbents.

Large enterprises don't buy point solutions; they invest in a long-term platform vision. To succeed, build an extensible platform from day one, but lead with a specific, high-value use case as the entry point. This foundational architecture cannot be retrofitted later.

Salesforce CEO Marc Benioff claims large language models (LLMs) are becoming commoditized infrastructure, analogous to disk drives. He believes the idea of a specific model providing a sustainable competitive advantage ('moat') has 'expired,' suggesting long-term value will shift to applications, proprietary data, and distribution.

Smaller software companies can't compete with giants like Salesforce or Adobe on an all-in-one basis. They must strategically embrace interoperability and multi-cloud models as a key differentiator. This appeals to customers seeking flexibility and avoiding lock-in to a single vendor's ecosystem.

In the future, it will be easier for businesses to build their own custom software (e.g., Salesforce) through prompting than to buy and configure an off-the-shelf solution. This shift towards "liquid software" will fundamentally challenge the one-size-fits-all SaaS model, especially for companies that currently rely on implementation partners.

The "horrific" user experience of Salesforce CPQ stems from a fundamental architecture problem. It was built for a simple "one seat, one license" world. The explosion of SKUs, consumption models, and complex discounting in modern SaaS has broken its underlying data model, creating a massive opportunity for AI-native challengers.