Alex Dunbrack

The $10M ARR myth that nobody wants to talk about → Stripe just dropped a bomb: Double the startups hit $10M ARR in 3 months in 2025 vs 2024. Everyone's celebrating. I'm concerned. Here's why this "achievement" might be the worst thing happening to the startup ecosystem: → Speed doesn't equal substance → Revenue doesn't mean profitability → Fast growth often masks fundamental problems The uncomfortable truth? Most of these "instant success" stories will be dead in 18 months. Why? Because hitting $10M ARR in 90 days creates a dangerous illusion: • You think you've figured out product-market fit (you haven't) • You scale before validating retention (customers churn) • You raise capital on vanity metrics (burn rate explodes) • You hire aggressively without culture foundation (team collapses) The data Stripe conveniently left out? Customer retention rates. Churn metrics. Actual profitability. I've seen this movie before. The 2021 "growth at all costs" era destroyed more companies than it created unicorns. AI is making it easier to acquire customers fast. But AI can't fix: - Poor unit economics - Weak value propositions - Non-existent customer support - Unsustainable business models The real question isn't "How fast can you hit $10M ARR?" It's "Can you still be in business 3 years after hitting $10M ARR?" Because sustainable growth > explosive growth. Every. Single. Time. Are we building the next generation of category-defining companies, or just creating a new bubble of revenue-rich, profit-poor startups destined to implode? What's your take – is ultra-fast ARR growth a feature or a bug of the AI era?

5

9,000 healthcare workers will lose their familiar AI tool in 2026. UCSF chose OpenAI over their own creation. The reasons reveal everything. UCSF just announced they're replacing their in-house AI platform Versa Chat with ChatGPT Enterprise. This isn't a small move. It affects 9,000 users across one of America's top medical centers. Why abandon your own technology? The answer shows where healthcare AI is heading: • Enhanced HIPAA compliance support • Access to GPT-5 technology • Enterprise-grade security safeguards • Broader functionality across clinical operations UCSF built Versa Chat for secure AI access. But even they recognized when to pivot. Sometimes the best solution isn't the one you created. The migration starts early 2026. UCSF's 40,000+ community members will gain access to cutting-edge AI while maintaining patient privacy. This shift signals a broader trend. Healthcare organizations are choosing proven enterprise solutions over custom builds. The focus has moved from building AI tools to implementing them effectively. For healthcare innovation, this could accelerate AI adoption across research, education, and patient care. What matters most? Getting the right tools to healthcare workers quickly and safely. How do you think this will impact AI adoption in your healthcare organization? #HealthcareAI #OpenAI #DigitalHealth 𝐒𝐨𝐮𝐫𝐜𝐞: https://lnkd.in/dSUD8XdG Karaka Sai chandra sekhar, Rupa Sri Borra

1

1 Comment

on ETA (across AI Roll-Ups funded by VCs, PE-style buy and build, HoldCos, traditional search funds, etc.) Context of my learnings: I've looked both at individual search fund deals from an investor perspective and larger platform ideas as founder. Over the past couple months, I've screened ca. 200 investment memorandums, did >10 site visits / superficial DDs, 2 deeper DDs, took 0 final leaps. I can't speak from the perspective of someone who's done it, but have talked to enough people to see patterns. Good news: The succession opportunity is real, the lag of technology adoption in SMBs is INSANE and there are some (!) high-quality SMBs with customer relationships so sticky, that it makes sense to continue something existing vs. building from scratch. I personally also believe many great operator talents exist, who have what it takes to lead an SMB, but might not be "0 to 1" founder types - and any form of entrepreneurship is net positive!! Bad news: Forget about risk-free returns and think through operational nightmare scenarios before you're stuck with them. Highly recommend the content of Alex Prokofjev / RollUpEurope, Christopher von Wedemeyer / Legacy Partners and Sahil Patwa / AI Roll-up Nexus to learn about common pitfalls. 1) Historical numbers suggest ca. 50% of search funds fail. Now adding ambitious add-on strategies, a wave of consulting / banking talents with no operator experience, and rising multiples.. I think this might trend up to 60-70% rather than down? Better than 90%+ failure risk of a startup, but no magic trick. 2) Thinking about the exit assumption, which is multiple arbitrage for most strategies: Which business really gets better by pasting together a couple of smaller ones, with different cultures, compensation practices, customer contracts...? Some do, because scale benefits can be powerful, but many don't. When speaking with people, who operated or acquired small cap platforms from leaders like Auctus / Ufenau / etc, it becomes clear that operational nightmares are the norm rather than the exception and the multiple arbitrage narrative could fade for many verticals. 3) Some VCs betting on deeper, longer-term tech value creation (...and maybe an IPO...in 20 years?) underwrite 15% to 40% EBITDA margin uplifts = relying less on multiple arbitrage, and maybe that's possible! But some of the "red hot" verticals are turning so competitive, with players funded near software multiples (have heard there are several already well-funded property managers raising again now), that sellers will realize they can push entry multiples up. VC-backed founders will feel pressure to put capital to work and will make mistakes. 4) More decentral HoldCos, with very high investment selection bars and less hands-on value creation, might bring the best risk-adjusted returns. However, it seems hard to attract brilliant young "builder" talent for those strategies and I'd personally always feel an itch to be more operationally involved.

293

27 Comments

CRV Security: Request for Startups I never know if this actually works for our friends over at YC but figured we'd try. Here's what we want to fund in 2026! 1. Golden Artifacts: Think Chainguard but more broad. Artifact attestation exists for open source. Almost nothing exists for internal software — especially the vibe-coded tooling now running in production. We want the company building cryptographic proof of secure software delivered from secure artifacts: who built it, how, and whether it was reviewed. If more things are being yeeted into the world via Claude Code (myself included), this feels like an issue. 2. MCP & Agentic Security: Agents are getting real credentials and taking real actions. The security posture of most orgs around this is basically zero. That changes fast. You'd never give an employee hardcoded API keys or write access to your email without supervision/trust. Why give it to agents? 3. AI Governance: Boards are asking CISOs to account for AI risk. CISOs have no good answer other than "Palo has a module" 4. Next-Gen Endpoint: CrowdStrike was built for a world of static binaries and human operators. AI workloads, cloud-native infra, and AI-assisted attackers need a new architecture. The category is ready to be reinvented. 5. Networking in the AI Era: Zero trust was designed for humans. What does network security look like when the entity requesting access is an agent? Nobody's really solved this. 6. Email Security + Next-Gen Phishing: LLMs have made spear phishing infinitely scalable. I've never truly understood why Abnormal and KnowBe4 aren't one company. Maybe this time it's different. 7. Frontier Security Lab: We'd back a credible, well-staffed lab focused entirely on red-teaming models and setting the evidentiary standard the industry needs as LLM built apps become the norm. 8. Dependency Security: That Actually Remediates Malicious and vulnerable dependencies are a top attack vector. The tooling is mostly noise — scanners that don't close the loop. The winner here ships fixes, not just alerts. 9. Critical Infrastructure Cyber: Data centers, satellites, power grids, undersea cables. The physical backbone of the internet is increasingly exposed and wildly under-defended. We have data centers in space, for God's sake. Surely we need better cyber for critical infrastructure? 10. PAM for the Modern Era Legacy: PAM was built for static roles, human users, on-prem directories. Cyberark was founded in 1999.....Agents, ephemeral workloads, and cloud-native infra have broken all of those assumptions. Is anyone rebuilding this from scratch? If you're building in any of these areas — or something we haven't thought of — reach out. james@crv.com

452

49 Comments

SF AI-Native Operator Takeaway #2: In AI-native PLG, the hard part isn’t conversion... it’s discovery. Many AI-native teams are still talking about PLG using a classic SaaS mental model, but based on operator conversations in SF, that model is starting to break down in fairly obvious ways. The biggest bottleneck right now isn’t conversion. It’s discovery. In traditional PLG, users generally understood the category before they ever signed up. The problem was obvious, the product’s value was legible from the homepage, and the “aha” moment tended to show up quickly in first use. In that world, PLG meant optimizing onboarding, reducing friction, and improving free-to-paid conversion because user intent already existed. AI changes that assumption. In AI-native products, users are often curious but unclear. They don’t yet know what’s possible, value depends heavily on workflow, context, data, and role, and the product can feel abstract until it’s applied directly to their job. As a result, many users stall not because the product isn’t valuable, but because they haven’t discovered how it fits into their world and how they can't live without it. This is the real distinction people kept coming back to. PLG conversion answers, “Is this worth paying for?” PLG discovery answers, “What problem does this solve for me, right now?” What’s working best in practice is less about funnel polish and more about clarity up front: role- or workflow-specific entry points, guided examples instead of blank states, and opinionated first actions that show users a concrete outcome before asking them to explore. This also explains a broader pattern across AI-native companies. Forward-deployed teams and services-heavy delivery aren’t just implementation tools; they’re discovery mechanisms. They translate abstract AI capability into concrete workflow value, observe real use cases users wouldn’t self-discover, and feed those learnings back into what eventually becomes productized. PLG isn’t going away, but in AI-native companies it’s being redefined. Self-serve no longer means self-explanatory. Education becomes part of the product, and discovery has to come before optimization. The teams making progress aren’t obsessing over conversion rates yet. They’re focused on whether users see themselves in the product, how quickly they reach a meaningful outcome, and whether the product helps users get to a meaningful outcome for themselves quickly, without too much guesswork. Bottom line: in AI, PLG is less about removing conversion friction early and much more about creating understanding first. Once they understand, they may be hooked. Tomorrow is my last SF AI operator takeaway focusing on everyone's favorite topic du jour: 996 work schedules.

15

1 Comment

VC pattern matching is broken. And that's your biggest opportunity. They're still looking for: → Stanford CS grads → Perfect pitch decks → 6-month roadmaps → $2M seed rounds → "Defensible" tech moats While solo builders are eating this up. I've watched this play out 50+ times: VC-backed startup: Still in beta after $500K Solo founder: $10K MRR with a laptop Why this creates massive opportunity: 1. VCs ignore "unsexy" problems → Dental inventory management → Restaurant scheduling 2. They underestimate non-technical founders → 15 years in the industry → Knows every pain point 3. They overvalue complexity → Simple tool, boring niche = They sleep → Same tool solving real problems = $30K MRR The tools that broke their model: → Lovable: Build without coding → Cursor: Refine with AI → Supabase: Backend that scales → ChatGPT: Validate before building Total cost: $100/month Time to first customer: Days While VCs chase the next AI unicorn: Plumbers are building plumber software. Dentists are building dental tools. Marketers are building marketing apps. And they're winning. Because they understand something VCs don't: The best person to solve an industry's problem is someone who's lived it. AI just gave them the tools. The window is wide open. But it won't stay that way. Once VCs figure this out, they'll flood these markets. Right now? They're still pattern matching for the wrong patterns. That's your advantage. Use it.

38

3 Comments