Interview, Fireside Chat
How Emergence Capital Bet Early on Zoom, Salesforce, Veeva—& Now Together.ai
- SaaS is predicted to evolve rather than disappear by 2030, with recurring per-seat models persisting but becoming integrated with AI cores and hyper-dependent on end markets; future app layer companies in 2025 will differ from traditional SaaS by emphasizing applied research and proprietary data harvesting.
- Infrastructure and application layers are expected to converge within two years, favoring nimble teams and high capital efficiency to build iconic businesses with limited funding, contrasting with the previous era of raising billions prior to IPO.
- The 2025 IPO market is anticipated to be weaker than hoped due to market run-ups, instability, and potential policy-driven interest rate increases or inflation, complicating timing for companies anchored to 2021 valuations.
- Interest rates are projected to remain stable or rise rather than decrease significantly in the near term, necessitating that software companies achieve high capital efficiency and excitement to justify historical valuation multiples.
- Significant capital allocation is forecasted for the $500 billion Stargate commitment, with an estimated $100 billion as risk equity, 25% for buildings and cooling, and 75% for chip acquisition, driven by national security imperatives against China rather than purely commercial logic.
- AI innovation faces a plateau with public data exhausted, shifting the focus to commercial data, private models kept within secure premises, and exclusive partnerships to access rich data assets to avoid a pricing race to the bottom.
- Emerging pricing models for 2025 will blend traditional per-seat pricing with usage gates, outcome-based, and volume-based structures, while new interaction modalities like voice and vision are expected to open entirely new software categories.
- Competitive advantage will rely on improving cost, speed, and quality, with a critical need for proprietary data moats; companies failing to secure such moats or demonstrate post-sales value in compressed sales cycles will struggle.
- Future workflows will persist as long as humans remain critical to advancement, potentially shifting from typing to neural links or chat bots, with cloud infrastructure often proving superior to single models for specific cost-reduction use cases.
- The firm plans to develop new investment theses every two to three years, maintain a "five-way marriage" partnership via quarterly external facilitators, and prioritize security, privacy, and trusted modality in voice platforms for complex interactions.
- Regulatory and legal frameworks remain unprepared for current model security needs, creating significant risks if business processes leak into public models, prompting a focus on controlled deployment and detecting malicious behavior.
- Investment criteria are shifting to include metrics where "engagement is revenue" and valuing post-sales experience, while smaller venture firms are expected to avoid scaling via secondaries that require huge portfolios.
- Unobvious opportunities are predicted to drive major wins in 2025, with high capital efficiency cases becoming replicable through tailored playbooks rather than exact traditional formulas.