David Sacks has a bold thesis:
“AI is following the exact same pattern we saw with SaaS 2.0.”
.png)
He’s not just making a prediction—he’s laying down a playbook for founders and VCs who want to surf the AI wave without wiping out.
In this post, I’ll break down:
SaaS 2.0 refers to the era from ~2010 to 2020, when:
This cycle produced giants: Zoom, Shopify, Datadog, Atlassian, Notion.
Craft Ventures was born during this time—and Sacks had a front-row seat.
Sacks sees the AI boom mirroring SaaS 2.0 almost identically:
PhaseSaaS 2.0AI BoomInfraAWS, APIsGPUs, models, vector DBsToolingZapier, SegmentLangChain, PineconeAppsNotion, ClickUpJasper, Adept, PerplexityPlatformsSalesforce, HubSpotTBD — OpenAI, maybe ChatGPT?
We’re currently in the apps phase, with early winners emerging.
If AI is repeating the SaaS playbook, then:
Your moat isn’t the model.
Your moat is distribution and usage velocity.
“Just like SaaS 2.0, AI-native startups win by going vertical, not broad.”
He’s betting on AI-native SaaS, not wrappers or infra clones.
Sacks is encouraging founders to:
The same playbook that worked for ClickUp and Figma can work for AI—but faster.
Sacks warns: “Don’t throw efficiency out the window.”
Even in AI, Craft uses:
AI doesn’t exempt you from fundamentals.
He’s bullish on:
Basically: AI that saves time, not just looks cool.
Here’s what Sacks calls out as red flags:
❌ API wrappers with no real IP
❌ No clear GTM strategy
❌ Products that rely on novelty, not retention
❌ Startups building on OpenAI with no hedge
Sound familiar?
It’s the same mistakes made during the SaaS gold rush.
Founders who:
✅ Solve a specific pain
✅ Layer AI into daily, high-frequency tasks
✅ Build an efficient GTM motion
✅ Understand distribution is king
Sacks says: “Don’t build a science project. Build a product that sells.”
He believes the AI stack is forming now, similar to the SaaS infrastructure wave in 2015–2018.
Here’s a rough parallel:
SaaS 2.0AI TodayAWS, GCPNVIDIA, AWS BedrockSegment, SnowflakeLangChain, PineconeZapier, RetoolLLM orchestration, AutoGPTHubSpot, Salesforce??? (still emerging)
We’re 2–3 years away from full-stack maturity.
Sacks is funding:
Notably, he’s not chasing flashy consumer AI.
He respects the tech—but warns founders:
“Don’t build your business on someone else’s API without a moat.”
If OpenAI launches a native version of your app tomorrow… do you survive?
✅ Own distribution (community, content, network effects)
✅ Layer in proprietary data
✅ Add workflow depth, not just UI
✅ Build trust over time
✅ Become irreplaceable in your user’s daily life
Sacks calls this the “Notion vs. copycat” test.
In 2021, SaaS was getting 100x ARR.
Now, early AI startups with zero revenue are raising at $100M+ pre.
Sacks warns founders:
1. What is SaaS 2.0?
The second wave of cloud software—focused on verticals, PLG, and modern infra.
2. How is AI like SaaS 2.0?
Same cycle: infra → tools → apps → platforms.
3. Does David Sacks invest in AI?
Yes—especially AI-native SaaS with traction and defensibility.
4. What’s his biggest concern?
Founders overfunding, overhiring, and building wrappers with no moat.
5. What AI metrics matter most to Craft Ventures?
Usage, retention, burn multiple, speed to value.
6. Does Sacks prefer custom models or API-based apps?
He prefers startups that control their stack and IP.
7. Is AI just hype?
No. But distribution, not model access, will define the winners.
8. How soon should AI founders monetize?
As early as possible. Don’t wait for scale—charge for value fast.
9. Will Craft lead rounds?
Yes—especially at Seed and Series A with strong PMF.
10. Is the AI bubble going to burst?
Valuations may correct—but real AI use cases will endure.
David Sacks isn’t just riding the AI wave.
He’s watching for the repeatable patterns from SaaS 2.0—and betting on the founders who see it too.
If you’re building in AI, don’t just follow the hype.
Follow the playbook that already created billion-dollar SaaS companies.
Subscribe to Capitaly.vc Substack (https://capitaly.substack.com/) to raise capital at the speed of AI.