Jason Lemkin on SaaS Investing, Churn and Seed Valuations

Jason Lemkin opens with strong claims about seed investing, growth and churn. In conversation with Harry Stebbings, he explains the assumptions behind those views and reflects on the investments that shaped them.

Lemkin separates his best investments by cash returned from those with the highest net asset value on paper. At the time of the interview, the two lists did not overlap; paper valuations are not cash distributions and may change.

He names Salesloft, Pipedrive and Logikcull as his top cash outcomes, describing sales of approximately $2.5 billion, $1.5 billion and $300 million respectively. He says his entities owned over 20% of Logikcull; these are his recollections, not audited fund returns.

He did not predict those three would lead by cash returned. In Salesloft, founder Kyle Porter and product leader Rob Forman showed exceptional commitment, he says. After investment, they abandoned an approximately $8 million revenue product to pursue a different direction.

Lemkin initially doubted that Salesloft and Outreach could both succeed in such a competitive market. Looking back, he prizes a complementary founding pair that can build and sell and is willing to persist; he describes this as his investment preference rather than a guarantee.

Early investors once had months to observe products before deciding, he says; compressed fundraising timelines now leave less time to assess people. His response is to meet the technical co-founder immediately, often on the second call, instead of waiting until the end of diligence.

He asks the CTO for a personal product demo and looks for enthusiasm, original insights and an ability to explain what is broken or missing. He values candor about competitors and technical limitations, though his assessment remains subjective.

Lemkin estimates that many early-stage CEOs he sees are stronger than their CTOs. A distinctive feature can help a young company reach its first customers, he says, but may mask a team that cannot iterate quickly enough to build a broader platform.

He says he would decline an otherwise attractive deal if the technical co-founder did not meet his standard. Illustrating dilution, he estimates a 12% seed stake might become 6% by exit, making a $1 billion sale worth $60 million to that position before other fund-level considerations.

Earlier in his career he interviewed broader management teams and once discovered a sales pipeline was weak despite promising headline metrics. For companies with ten or fewer employees, he now focuses more on the CEO and CTO than on the quality of every early hire.

To develop a benchmark for technical leaders, he suggests meeting roughly 15–20 CTOs, asking open questions about product demos and feature gaps, and listening for insight. He recalls learning from a trusted engineering colleague rather than relying on job titles alone.

He also examines product quality: slow dashboards or broken demonstrations at an early revenue milestone concern him because more users and workflows may compound those problems. This is his diligence heuristic, not proof that every imperfect product will fail.

Stebbings dislikes competitive markets. Lemkin shares the discomfort but says he regrets automatically passing on exceptional founders because a market already has incumbents. Large markets often attract many competitors, and strong teams can benefit from the education those competitors provide.

Conversely, a market with little modern competition may still conceal an underpowered product team. Lemkin wants software to keep improving over years rather than relying indefinitely on a few clever features.

Pipedrive sold to Vista for approximately $1.5 billion, with around $100 million in annual recurring revenue at the time, according to Lemkin. He says his combined ownership was close to 10% and reflects that five co-founders involved in decisions can be difficult to coordinate.

He recalls investing when Pipedrive was already growing quickly and stood out among many small-business CRM products. In his view its product later changed too slowly while HubSpot expanded into CRM. His comparison reflects his perspective as an early investor, not an independently established competitive analysis.

Lemkin says he generally wants to sell a significant position when founders leave, because he values the competitive agility of founder-led companies. He acknowledges that others disagree and that professional leaders can be necessary; his rule is a personal investing preference.

Stebbings predicts that some private-equity buyers may lose money on highly priced software acquisitions. Lemkin says he does not know the result and asks whether limited partners will tolerate weak 2021-vintage deals. Their discussion is speculation, not evidence of losses on named transactions.

Exit timing matters, Lemkin says, pointing to the high prices available in 2021. He recalls Salesloft founder Porter pushing for the sale despite mixed investor views; Lemkin now says he is less inclined to talk a founder out of a sale they believe is right.

HubSpot’s CRM expansion illustrates the value of sustained founder commitment, he argues. Although it initially served marketers rather than sales teams, HubSpot invested in an adjacent product for years. He similarly cites Datadog’s expansion as a reason not to assume larger founder-led competitors will move slowly.

Lemkin uses roughly 10% share of a company’s core ideal customer segment as a signal to plan a second act. Growth may start becoming harder before headline results show it; expansion could mean a new product, a larger customer segment or another geography. The threshold is his rule of thumb, not a universal law.

He says early Pipedrive investors sometimes imagined it as the next Salesforce even though the founders described a self-service small-business product. His lesson is to listen to a founder’s actual ambition rather than imposing a convenient investment narrative.

One of his most disappointing outcomes still returned five times his investment when a company sold for $100 million, he says. He believes poor executive hires increased spending while sales weakened, and that a founder tolerated the situation too long. The foregone larger outcome is hypothetical.

He describes another unnamed company where he ultimately lost $5 million. Lemkin alleges that its CEO misrepresented some financials and says he nevertheless wrote a third check as growth accelerated in 2021. The company is unnamed and the account is his, not an independently adjudicated finding.

His lesson is to apply diligence to follow-on checks rather than relying on top-line growth. He has moved financial checks earlier and asks his accounting firm to compare financial information with bank statements, seeking reasonable accuracy and signs of manipulation rather than perfect startup accounting.

He regards customer references as a significant request of a founder and prefers to make them when he is already seriously interested. At around $1 million in revenue, strong growth and low churn lead him to expect broadly satisfied customers, whose feedback can confirm or challenge that expectation.

On growth, Lemkin says the strongest investments in his experience can move from $1 million to $10 million in roughly five quarters or less. He describes 8% monthly growth around $1 million in revenue as a personal bar and double-digit monthly growth as exceptional; these are venture-screening heuristics, not requirements for every viable business.

Asked whether extraordinary AI startups have changed expectations, he says fast-growing outliers are attracting concentrated capital. In his portfolio, he sees existing investors financing winners aggressively, sometimes at inflated prices. This is an observation from his own investments, not a market-wide measurement.

He worries that insiders can justify a high new valuation by averaging it against their earlier low entry price, leaving a company overfunded. Stebbings raises structured or flat rounds used to avoid valuation cuts; Lemkin acknowledges limited visibility and distinguishes ordinary bridge financing from repeated support for a struggling business.

Stebbings asks about slower-growing public software companies. Lemkin resists calling them zombies: they may be profitable or become targets for private equity. He cites Model N and discusses software combinations, but says he does not know whether individual buyouts will succeed.

He says a mature company with high operating margins can remain a cash-generating business even if its growth is less exciting. Improved growth or public valuation multiples could make some private-equity deals attractive, although he does not predict which ones will work.

For enterprise software at roughly $1 million in revenue, Lemkin looks for net revenue retention above 110% and says retention below 100% would concern him. He frames this around his own experience and venture criteria, not a universal statistical cutoff.

For very small business customers, he describes monthly customer churn of around 3–4% as common in his experience, and finds higher rates troubling for a recurring-revenue model. He notes uncertainty about historical HubSpot figures and stresses that a company ultimately needs a path toward strong net retention.

When monthly figures vary sharply, he likes averaging the latest four months of churn, growth and burn as a planning heuristic. He considers it useful for forecasting, but the conversation does not establish that it will predict every startup accurately.

High small-business churn can be survivable with extremely low acquisition costs and efficient operations, he says; a costly sales-led motion makes the economics harder. He wants evidence that founders can improve retention or move upmarket, rather than masking the problem with funding.

Lemkin recalls an investment with roughly 2% monthly churn whose team struggled to solve the issue even while capital supported growth. He says small businesses can cancel quickly, leaving less time to correct mistakes than a multiyear enterprise contract might allow.

Burn cannot be assessed in isolation, he argues. A burn ratio that looks efficient with strong margins and retention may be inadequate for a business with weaker retention, hardware costs or uncertain future financing. Customer payback periods matter especially for very small businesses.

He cautions against transferring enterprise sales and marketing playbooks directly to a small-business model: customer lifetime, spending and acquisition economics differ. Asked about conservative projections, he suggests giving a promising first-time founder the chance to explain or correct a model before passing.

Stebbings asks how to distinguish sustainable AI revenue from experimental spending. Lemkin says the answer is not yet known. He is more comfortable underwriting rapid growth with positive cash flow than a similarly fast-growing company burning enormous sums, though he admits the latter can also succeed.

The investors agree that an excellent founder does not automatically make any price sensible. Lemkin worries that high seed valuations and small ownership stakes make fund returns difficult: a small fund must choose between concentration, tiny positions and earlier, riskier deals. This is his assessment of the market, shaped by his own fund constraints.

He argues that Y Combinator’s incentives favor getting many companies funded quickly, rather than ensuring every seed fund can obtain its preferred ownership. He treats this as a structural mismatch, not an obligation for YC to redesign its process for investors.

Lemkin recounts investing in RevenueCat when it had only about $30 in monthly revenue. He says he initially confused subscription volume managed on the platform with its own revenue, taking more early risk than intended. He estimates the pre-YC valuation at $7–8 million and his initial stake at about 10%.

What gave him confidence was the founders’ knowledge of the market and long-term commitment. He praises RevenueCat’s developer product while acknowledging that sales-led growth developed more slowly than it might have; his favorable description comes from an investor in the company.

He contrasts that confidence with an unnamed loss where founders, in his account, cut sales and marketing after growth became difficult despite money in the bank. He says a seemingly favorable price cannot turn a weak investment into a strong one and regrets treating an end-of-fund deal differently.

He also wishes he had recycled cash from an early exit into the portfolio instead of distributing it, believing additional investment could have improved fund returns. He acknowledges that fund recycling is complex; the additional return he imagines is hypothetical.

His limited partners tell him he invests too slowly, and he agrees it is a concern. He believes his existing investments may meet his financial goals but hopes to raise more funds and back additional companies capable of returning a fund; these are aspirations, not realized results.

In quick fire, he names ServiceTitan’s Ara Mahdessian and Klaviyo’s Andrew Bialecki as underrated SaaS leaders. He thinks Klaviyo is undervalued relative to some public peers and discusses possible future growth; his valuation judgment and forecasts are not established facts.

Asked about Anaplan and private-equity ownership, Lemkin says he does not know what will happen. He notes that entrenched enterprise products can be hard to replace and that even moderate growth plus strong margins may support substantial value, despite criticism of innovation.

He says a small number of supportive limited partners and affiliates account for a large share of his funds. Although concentrated funding brings risks, he values the ability to work with backers who understand his approach; his experience is not necessarily representative of other managers.

Lemkin advises individuals to consider illiquidity and the long time needed to realize venture returns. Asked to select funds for an institution, he says historical records can obscure partner departures and changing incentives, so the people making future decisions matter.

He speculates about why investment-firm partners might step back after a huge success and illustrates hypothetical carry calculations using Datadog and other examples. He explicitly says he could be wrong about any particular firm; these figures and motives are not verified compensation disclosures.

He closes by saying that large past successes do not guarantee enthusiasm for another long and difficult fund cycle. For Lemkin, the important question is whether an investor still has the drive to do the work; he thanks Stebbings and looks forward to meeting in London.