So Rory from Scale, Harry from 20VC and me were back this week on another deep dive on just what’s happening in B2B today:
- The latest mega raise from Thoma Bravo
- If OpenAI is eating B2B
- The MCP threat to SaaS
- When to Hold and When to Fold as a VC
and much more!
It started off admittedly awkward with a guest VC who was pretty harsh on SaaS and B2B. Skip that part if you want. I don’t approve of it and didn’t enjoy it. But the parts after were good. e.g. on the AI Slow Roll:
The Brutal Truth About Today’s SaaS Market: From DPI vs TVPI Wars to the AI Slow Roll Killing B2B Companies
A deep dive into the most pressing issues facing SaaS founders and investors in 2025, based on insights from top VCs and operators
The venture capital world is having an identity crisis. And if you’re a B2B SaaS founder, the ripple effects are coming for you whether you’re ready or not.
In a recent conversation between some of Silicon Valley’s most candid VCs, a fascinating picture emerged of an industry in transition—one where the old playbooks are breaking down and new realities are forcing uncomfortable reckonings.
The Great DPI vs TVPI Debate: What Actually Matters for SaaS Success?
The conversation started with a provocative take from Chamath Palihapitiya: “TVPI is bullshit. You can’t eat IRR. You can only eat net DPI.”
For those not versed in VC speak, this cuts to the heart of how we measure success. TVPI (Total Value to Paid-In capital) includes paper markups—those impressive valuations that look great in quarterly reports. DPI (Distributions to Paid-In capital) is cold, hard cash that’s actually been returned to investors.
The reality check: “There’s two very different games called venture capital. One game is actually making people money—finding companies early, making the right bets, paying the right prices, and selling. That’s a DPI game. There’s also an asset gathering game, and frankly, as a business, the asset gathering game is actually better if you’re just in it to make money.”
What this means for B2B founders: Your investors are increasingly focused on real exits, not just pretty valuations. The days of celebrating another markup round without thinking about liquidity are numbered. Smart founders are already having conversations about secondary sales and thinking strategically about when to take money off the table.
Tom O’Bravo just raised a record $34 billion fund with $30 billion in distributions last year—not paper markups, but actual cash returned to LPs. That’s the new gold standard.
The Hollowing Out of Mid-Tier VC: A Crisis for Series A SaaS Companies
Perhaps more concerning for SaaS founders is what’s happening to the venture ecosystem itself. Mid-tier VC funds (roughly $200-500 million in size) are getting “annihilated,” according to recent SVB analysis.
The new reality is binary: You have massive $5-10 billion funds that can write $50+ million checks, and you have smaller $200 million funds focused on seed and early-stage deals. The middle is disappearing.
Why this matters for Series A SaaS companies: If you’re looking to raise a $20-30 million Series A (which is now considered normal, not large), your options are increasingly limited to mega-funds. These funds are writing 40-60 checks of this size annually, creating a much more competitive and less personalized fundraising environment.
As Rory noted: “If there’s only $200 million funds and billion dollar funds, then logically, those 20, 30, 40, 50 million dollar checks are going to come from only 10 names who are writing 40 of them a year, not eight like us.”
The founder implication: You better get to know decision-makers at mega-funds before demo day. The personalized, relationship-driven Series A process is becoming a luxury good.
The $600 Billion AI Infrastructure Question: Where’s the Revenue?
Here’s a staggering number: The big six tech companies spent $212 billion on CapEx in the last year, mostly on AI infrastructure. That’s unprecedented in internet history.
The disconnect is real: While infrastructure spending has exploded, application-level revenue is still catching up. OpenAI is projected to hit $25-30 billion in revenue next year, and Anthropic crossed $3 billion (up from $1 billion just five months ago), but relative to the infrastructure spend, we’re looking at “$585 billion in losses” against $10-15 billion in revenue.
For B2B founders, this creates both opportunity and pressure:
- Opportunity: The infrastructure is being built for you. As one investor put it, “Intelligence is available at a price that’s declining by log orders of magnitude every year.”
- Pressure: Public company investors are “mean VCs on steroids”—they can turn “on a dime from ‘I can’t believe you’re not spending more’ to ‘what the frick do you mean you’re spending so much money?'”
The takeaway: Token costs have collapsed 99.7% in two years. If you’re a B2B SaaS company and your VP of Engineering is still saying AI integration is “too expensive,” it’s time for a serious conversation about capabilities and leadership.
The AI Slow Roll: The #1 Killer of B2B SaaS Companies
This brings us to what might be the most important insight for SaaS founders: the “AI slow roll” is becoming an existential threat.
What the AI slow roll looks like:
- “We’re talking about AI implementation in Q4”
- “We’ll do a limited release next year”
- “We’re still evaluating the best approach”
Why this is fatal: As Varun from Windsurf (who’s had three different companies in 18 months) put it to Harry: “Startups beat incumbents because of existential dread. If you are in a startup and you don’t ship great products that convert to sales, you lose.”
The most successful SaaS leaders are embracing this dread. Aaron Levie at Box and Yamini Rangan at HubSpot both said the same thing at SaaStr 2025: “We’re so excited and we’re scared.”
What existential dread looks like in practice:
- Hackathons every weekend
- AI voice agents shipping this week, not next quarter
- Company-wide all-hands focused on AI integration
- Leadership visibly shaking about competitive threats
If this isn’t your reality, you’re probably going to fail.
The Coming User Interface Revolution: Why Your SaaS App Might Become Invisible
Perhaps the most profound shift coming is how the next generation will interact with software. Mary Meeker’s latest AI report highlighted that 32% of the world is just coming online, and they’ll be “AI-first” users.
This generation will not:
- Use traditional SaaS interfaces
- Understand concepts like “files”
- Navigate through complex menu systems
- Think in terms of “applications”
Instead, they’ll expect:
- Voice-driven interactions
- Agent-based workflows
- Natural language commands
- Seamless cross-platform experiences
The MCP (Model Context Protocol) threat: This is where things get existential for SaaS companies. MCP allows AI systems to interact directly with your application’s data without users ever touching your interface.
A real example: Jason shared the story of MangoMint, a SaaS platform for spas and medical offices. The CEO realized that with MCP, customers could book appointments through ChatGPT or Claude without ever knowing which booking system the business uses. “I could become a pipe overnight,” he said.
The implication: If users can accomplish their goals through AI agents that abstract away your interface, your entire value proposition needs to evolve. You need to become the system that helps make decisions, not just the system that tracks what happened.
Market Share Math: When to Hold, When to Fold
For SaaS companies approaching maturity, there’s a crucial mathematical reality around market share that most founders ignore.
The 15% rule: When examining companies like Chime and Revolut, several investors noted that once you hit about 15% market share in your true TAM, growth dynamics fundamentally change. CAC inevitably increases, competitive pressure intensifies, and the “easy wins” disappear.
This creates a strategic decision point: Do you optimize for infinity growth, or do you recognize when you’ve built a great business that should be monetized?
Recent examples:
- Chime’s upcoming IPO at ~$12-14 billion (down from $25 billion last round)
- HubSpot trading around all-time highs while facing AI disruption threats
- Salesforce acquiring Informatica for $8 billion to stay relevant
The key insight: There’s a difference between companies that “matter” (platform-shifting, $100+ billion market cap potential) and companies that are “perfectly good businesses” ($5-15 billion market cap, consistent returns).
Both can make investors and founders wealthy, but the strategies for optimizing each are different.
What Actually Matters: The Uncomfortable Truth About Impact
The conversation took a philosophical turn when discussing what companies actually “matter” in the grand scheme of technological progress. Sam Lessin argues with Jason and the group about how meaninful SaaS is.
The harsh reality: Most companies, even successful ones, don’t fundamentally change the world. Lessin argued Box, despite being a well-run public company worth $5+ billion, probably doesn’t “matter” in the sense of being paradigm-shifting.
But here’s the nuance: Not mattering in a historical sense doesn’t mean not being valuable. As one investor put it: “If the $14 billion market cap was in my bank account, I would think it was a very important company indeed.”
For SaaS founders, this creates two strategic paths:
- Build for infinity: Go after platform-shifting opportunities where you could fundamentally change how an industry works (think Salesforce in the early 2000s)
- Build for excellence: Create a phenomenal business in a defined market, optimize for profitability and exits, and don’t apologize for being “just” a great business
The mistake: Trying to split the difference. Either embrace the infinity bet with appropriate risk tolerance and capital requirements, or build a capital-efficient business optimized for strong returns.
The IPO Market Reality Check: Normal Service Resumed
Despite all the doom and gloom about AI disruption and market dynamics, there’s actually good news on the liquidity front.
Recent IPO activity:
- Chime filing for public offering
- Circle IPO at $8 billion valuation
- Grow (India) filing for IPO
- Multiple other B2B companies going public
The key insight: “Price clears all markets.” Companies that couldn’t go public at inflated 2021 valuations are finding realistic pricing that works for public investors.
What this means for SaaS founders: The IPO window is open if you have real business fundamentals. Don’t wait for the “perfect” market—focus on building a business that can justify its valuation at any point in the cycle.
Database Wars and AI Infrastructure: Where the Real Money Is
While everyone focuses on application-layer AI companies, some of the smartest money is flowing to infrastructure plays.
Recent moves:
- Snowflake acquiring Crunchy Data (Postgres) for $250 million
- Databricks buying Neon for $1 billion
- ClickHouse raising at $6 billion valuation
The pattern: Data infrastructure companies are becoming AI infrastructure companies. The ability to efficiently store, process, and serve data for AI workloads is becoming a massive competitive advantage.
For SaaS founders: If your application generates significant data, think strategically about how that data becomes more valuable in an AI-driven world. Data moats are becoming AI moats.
YC Valuations and the New Math of Seed Investing
Y Combinator startups are now raising at $50-60 million post-money valuations with minimal revenue. This represents a fundamental shift in how early-stage companies are valued.
The new dynamics:
- 70% of YC batch companies have AI components
- $50-60 million is becoming the new standard for AI startups
- Series A rounds need to be at $150-200 million to justify the risk/return
For founders: This creates both opportunity (more capital available) and pressure (higher expectations, lower ownership for investors, need to execute at higher velocity).
The math problem: With VCs owning smaller percentages of companies, fund returns become more dependent on picking absolute winners rather than a portfolio approach.
Action Items for B2B Founders
Based on these insights, here’s what successful SaaS founders should be doing right now:
Immediate (Next 30 Days)
- Audit your AI integration timeline: If it’s longer than Q1 2025, you’re moving too slowly
- Assess your Series A readiness: Can you justify a $150+ million valuation? If not, bootstrap longer or raise a smaller round
- Review your secondary market options: Start building relationships with funds that buy from early investors
Medium-term (Next 6 Months)
- Redesign workflows for AI-first users: Your 2026 customers won’t use software like your 2024 customers
- Build MCP compatibility: Make your application data accessible to AI agents
- Stress-test your market share math: Are you approaching the 15% inflection point?
Strategic (Next 12 Months)
- Define your infinity vs. excellence strategy: Are you building for $100+ billion potential or optimizing for a $5-15 billion exit?
- Develop your data moats: How does your data become more valuable in an AI world?
- Plan your liquidity strategy: Secondary sales, strategic acquisition, or IPO pathway
The Bottom Line
The SaaS world is experiencing the most significant platform shift since the move to cloud. Companies that adapt quickly and embrace the uncomfortable realities of AI-first development will thrive. Those that don’t will become cautionary tales.
The good news? The infrastructure is being built for you, capital is available for strong businesses, and the IPO market is functional. The bad news? The competition is intensifying, user expectations are evolving rapidly, and there’s no time for the luxury of gradual change.
As one investor put it: “Every company’s market share is up for grabs when there’s a platform shift. The very biggest ones don’t go to zero—they just slow down. But if you’re running one of these companies and you’re not afraid and you’re not 110% focused, you’re almost certainly going to fail.”
The question isn’t whether AI will disrupt your SaaS business. The question is whether you’ll be leading that disruption or becoming a victim of it.
The choice, as always, is yours. But the window for making that choice is closing fast.
