Is SaaS Dead? 8 AI Startup Truths from Anthropic and Cursor
AI Startup Insights from SaaStr Annual Conference
From the SaaS Playbook to the AI Era
In the 2010s, the SaaS growth story was straightforward:
standardized products, clear sales processes, and predictable subscription revenue created a near‑replicable machine.
Artificial intelligence has upended that simplicity — reshuffling technology, delivery, product design, and commercialization.
Key differences in the AI era:
- Compute cost replaces code cost as the primary production factor.
- Gross margins are lower, even when revenue growth is exceptional.
- Traditional KPIs are obsolete, but new evaluation frameworks aren’t standardized yet.
At this year’s SaaStr Annual Conference, founders from Anthropic, Cursor, and fal explored how AI entrepreneurship is redefining economics and strategy.
---
8 Key Insights for Building AI Companies
01 — Old SaaS Metrics Don’t Apply
In traditional SaaS, new users = high margin.
In AI-native companies, every user consumes:
- GPU compute
- Electricity
- Model inference time
Margins of 40–50% are now common, versus 80–90% in SaaS.
Founders must rethink “healthy growth” by emphasizing unit economics and compute efficiency over legacy metrics like the “Rule of 40.”
---
02 — COGS is the New CAC
> “Cost of Goods Sold (COGS) is the new Customer Acquisition Cost (CAC).” — Talia
In SaaS, CAC was the key constraint.
In AI, compute cost is the bottleneck.
Teams must drive adoption via:
- Product quality
- Virality
- Community
Formula for success: High cost to serve + Low cost to acquire → Strong product stickiness.
---
03 — Price for Outcomes
AI monetization is shifting from seat-based subscriptions to:
- Usage-based pricing
- Outcome-based pricing
Why?
Each model run consumes real resources — GPUs, electricity, cooling.
Main Models Include:
- Pay-per-use: API calls, tokens, tasks.
- Outcome-based: Pay only for measurable results.
- Hybrid: Base subscription + usage tiers.
> “If AI can double developer productivity, pricing should match that impact.” — Jacob, Cursor
Key takeaway: Align revenue directly with delivered value — not just access.
---
04 — Shadow Targets for AI GTM Teams
Predicting AI adoption cycles is tricky.
Anthropic uses “shadow targets” (goals rooted in feedback and mission, not forecasts) for sales.
GTM adaptation:
- Small, technical teams
- Salespeople who code
- Rapid learning loops
- Process flexibility
---
05 — AI Embedded in Daily Operations
Examples of internal AI use:
- Anthropic: Claude-powered Slack assistant for onboarding & knowledge search.
- Cursor: Asynchronous “background agents” to assist developers.
- fal: Compute fellowship program to test talent before hiring.
Outcome: Leaner, faster, smarter teams.
---
06 — Commit to Category Leadership
fal focused solely on generative media inference — avoiding dilution by serving all models/modalities.
Lesson:
Clear category definition → Faster adoption → Stronger market position.
---
07 — Growth Through Collaboration
Co-opetition is emerging in AI.
Example: Anthropic + Cursor
- Anthropic’s Claude supports Cursor’s coding assistant.
- Model improvements benefit both sides.
Ecosystem mindset beats zero-sum competition.
---
08 — New North Star Metrics
In AI, the most important signals aren’t just ARR or margins — they’re usage, customer love, and leverage.
Track:
- Usage trends
- Internal NPS (does your own team love the product?)
- Logo diversity
- Share of wallet
> “Revenue lags user growth. User growth lags product quality.” — Jacob
Bottom line: AI products should be indispensable and beloved — more “colleague” than tool.
---
Practical Tools for AI Founders
Platforms like AiToEarn官网 help balance value creation with multi-channel reach:
- AI-powered content generation
- Simultaneous publishing to Douyin, Kwai, WeChat, Bilibili, Rednote, Facebook, Instagram, LinkedIn, Threads, YouTube, Pinterest, X
- Performance analytics & AI model ranking
Such ecosystems enable efficient scaling while staying focused on measurable customer outcomes.
---
Final Takeaway:
AI entrepreneurship rewards focus, value-based pricing, collaboration, and customer engagement over old SaaS metrics. Founders who master compute economics and tightly align their revenue model with delivered value will shape the next era of software.
