Most important take away

Seed investing is systemically broken because too much capital is chasing too few companies that can sustain triple-digit growth, and entry prices (often $25M pre at YC) make it nearly impossible to own enough to return a fund after dilution. The bar for a great SaaS investment is unforgiving: $1M to $10M ARR in five quarters or less, “triple-triple-double-double” growth, sub-3-4% monthly SMB churn, 110%+ enterprise NRR, and a CEO-CTO pair with insane commitment - anything less rarely produces venture-scale returns.

Summary

Jason Lemkin walks through his three best and three worst SaaS investments to extract actionable lessons. His best cash outcomes (SalesLoft at $2.5B, Pipedrive at $1.5B, and Logikcull at $300M) taught him that the binary CEO-CTO pair with relentless commitment matters more than any other factor. Because diligence windows have collapsed from months to a single hour, he now front-loads a Zoom with the CTO as the second call - having the CTO run the demo, asking what frustrates them, and asking for the top feature gaps. Great CTOs answer those questions in 60 seconds with hyper-transparent specifics; mediocre ones hedge. Lemkin says he can spot world-class technical talent within ten minutes, and slow software at $1M ARR almost never gets better at $10M.

Lemkin’s career advice for operators centers on commitment and pairing: build with a co-founder more committed than you are, and own technical excellence even at $1M revenue because the load only goes up. For founders crossing 10% market share in their core ICP, start the second act early - either a new product (Parker Conrad style), broader ICP, or geographic expansion - because growth slows after that threshold and you need runway to react calmly. The HubSpot vs Pipedrive case study shows that founder-led companies (HubSpot, Datadog with Olivier) can ruthlessly expand into adjacent categories and win even when the original buyer profile differs - so be wary when a founder-led incumbent targets your space. Conversely, when founders leave, sell - Lemkin would have liquidated Pipedrive when the founders departed, because professional CEOs lose the competitive agility that only founders preserve.

On benchmarks: at $1M ARR he wants 8-10% month-over-month growth (he regrets every time he lowered this bar), 110%+ NRR in enterprise (non-negotiable - if it’s not triple-digit at $1M, something is fundamentally broken), and at most 3-4% monthly churn for SMB. Anything higher than that and “it’s not even software anymore” - the unit economics of SaaS break without 100% NRR. He uses an L4M (Last 4 Months) average for growth, churn, and burn as a highly predictive forecast for the next 8-9 months. For SMB plays, the product must be self-serve/PLG by necessity because users will cancel at the next credit-card statement.

On worst deals, his biggest lesson is that follow-on checks deserve the same diligence as initial checks - his $5M loss came from a reflexive third check during 2021 when he should have stopped at the second. He also regrets letting end-of-fund psychology (“we’re already in carry, just recycle”) drive sloppier decisions. He now front-loads financial diligence (bank statements via his auditor) to screen for “bullshit” early, treats customer references as confirmation rather than tiebreaker, and refuses to invest when founders bring enterprise/mid-market operators into SMB businesses - the toolkits don’t transfer. Don’t take “good deals” - good deals don’t make great investments.

On the macro: insider capital is flooding the few outliers growing at triple digits, inflating valuations to levels that don’t make sense for new entrants. Many 2021 PE buyouts (Zendesk, Anaplan, potentially Pipedrive) likely won’t return money, but LPs are giving Mulligans. Zombie public companies (Dropbox, Box, Twilio) approaching 40% operating margins may still be decent investments if multiples re-rate 20%. Lemkin believes Klaviyo and ServiceTitan are the most underrated SaaS companies, and that Klaviyo’s 6x revenue multiple is anomalously low - if it traded at 20x like it should, the rest of the venture market would make sense.

For individual investors: don’t invest in venture funds personally - put it in VTI/S&P 500. Three to five committed LPs who let you operate like you operate with VCs is the sustainable model. And recycle aggressively - he wishes he had recycled the $100M exit instead of distributing it.

Chapter Summaries

The new diligence playbook (CTO first): Lemkin’s process has inverted. The second call is now with the CTO, where they demo their own product, name their proudest feature, and surface their top feature gaps. Great CTOs are hyper-transparent and answer in 60 seconds; mediocre ones hedge. Slow software or broken demos at $1M ARR are an immediate pass.

Best by cash - SalesLoft ($2.5B): The lesson is binary co-founder commitment. SalesLoft survived dumping an $8M product to near-zero revenue because Kyle and the CPO never wavered. Competitive markets (vs Outreach) can both have winners if engineering velocity compounds.

Best by cash - Pipedrive ($1.5B): Five co-founders is too many; product stagnated; HubSpot’s founder-led expansion into CRM eventually owned the category. Lesson: when founders leave, sell. Founder-led incumbents (HubSpot, Datadog) will decimate adjacent categories patiently over years.

Expanding the second act at 10% market share: As you approach 10% share of your core ICP, growth slows. Founders need a second act ready - new product, broader ICP, or geo expansion - so they can implement calmly rather than panic.

Worst deals - the $5M loss and the $5X “loss”: The $5M loss came from reflexively writing a third follow-on check in 2021 without diligence. The 5x return (sold for $100M during lockdown) was a loss in opportunity cost - the founder hired a bad CEO who hired 20 bad reps, halving sales while burn exploded. Stubborn founders who don’t fire fast enough kill outcomes.

Diligence reordered: Financial diligence (bank statements via auditor) now comes first to screen for “bullshit.” Customer references are confirmation, not tiebreakers. Don’t burn founders’ customer relationships unless you’re 90% sure you’ll invest.

Benchmarks - growth, churn, NRR: $1M ARR needs 8-10% MoM growth; enterprise needs 110%+ NRR (non-negotiable); SMB tolerates up to 3-4% monthly churn but not more. L4M averaging is highly predictive. Burn ratio matters less than runway-to-next-round and gross margin reality.

SMB vs enterprise: SMB software must be better than enterprise because there’s no contractual stickiness - users cancel at the credit-card statement. Don’t put service-now-pedigreed operators into SMB businesses; the playbooks are incompatible.

AI hype and capital flood: Don’t over-analyze AI revenue durability - that’s venture. Worry about burn rate. Insider capital floods the few triple-digit growers, inflating prices and reducing diligence. Low ownership stakes are the systemic risk LPs will quietly regret.

Seed is broken: With YC priced at $25M pre and 150 companies per batch, $50-100M seed funds can’t write checks large enough to own 10% post-dilution. YC’s model isn’t designed to make seed VCs happy.

Revenue Cat - what wins: Founders committed to a 20-year mission in a domain they know cold. PLG-only sales delays were forgiven because the team and conviction were unwavering.

Macro - PE, zombies, public markets: 2021 PE deals likely won’t return money but LPs will give Mulligans. Zombie public SaaS at 40% margins is investable if multiples re-rate 20%. Klaviyo and ServiceTitan are underrated; Klaviyo should be a $20B company.

LPs and recycling: Find 3-5 LPs who treat you like a partner. Recycle aggressively - distributed exits early in a fund’s life are missed compounding. Individuals should not invest in venture funds at all; use VTI.

Quick Fire: Most underrated SaaS founders - Ara from ServiceTitan and Andrew from Klaviyo. Most overpriced - the gap between Klaviyo (6x) and Atlassian (12x at 20% growth) makes no sense. Anaplan-style PE-owned non-innovators may still hit rule of 40 in the teens for a decade.