Tomasz Tunguz on GTM Leverage, AI Margins, and Why SaaS Founders Have Never Had It Better

TL;DR

In this episode of Second Acts, Chargebee CEO Krish Subramanian speaks with Tomasz Tunguz, founder and General Partner of Theory Ventures, about

• The three infinite learning curves of venture
• The limits of advice
• Why writing daily sharpened his thinking
• And his conviction that distribution is the most compounding force in SaaS

Tunguz also discusses why pricing is always in flux, why AI startups should sometimes embrace uneconomic growth — investing ahead of profitability to capture market share early; what Looker taught him about ‘extreme PMF’; and why he believes today is the best moment in history to build a SaaS business.

A walk along the Embarcadero

The seed of Tunguz’s venture career was planted on a weekend walk with his wife along San Francisco’s Embarcadero. He overheard an angel investor advising a founder.

“It blew my mind that this was an industry, this was a job,” he recalls.

That curiosity eventually led him into venture capital, and nearly two decades later, to founding Theory Ventures, a $680M fund backed by some of the most respected names in enterprise software.

Venture’s three infinite learning curves

For Tunguz, venture’s appeal lies in what he calls its three infinite learning curves. The first is obvious: staying on the bleeding edge of technology.

The second is subtler: evolving alongside founders as their companies scale through hypergrowth. The third emerged more recently: the venture industry itself is shifting, expanding toward a $500B asset class by the end of 2028. Each curve requires constant adaptation.

“Advice is a tricky thing.”

That dynamism also shapes how he counsels founders. “Advice is one person’s experience generalized,” Tunguz cautions. In rapidly changing contexts like AI, even proven playbooks require fresh interpretation. Instead, he believes effective board members contribute by asking simplifying questions and offering frameworks at the right time.

One of his favorites, borrowed from Hanabi Capital’s Mike Volpi: “Is this company product-limited or go-to-market-limited?” The answer, he says, often swings back and forth as markets shift.

Dalio, Munger, and effective board members

Great board members, Tunguz argues, don’t try to operate companies themselves. They clarify complexity with frameworks, much like Charlie Munger’s Poor Charlie’s Almanack or Ray Dalio’s Principles. “There are probably one to two priorities that really matter for a business at any given time. Exceptional board members help isolate those and provide a mental model for how to manage them,” he says.

Situation, Complication, Question, Answer

That same love of frameworks informs Tunguz’s writing discipline. For over 15 years, he has published essays almost daily, making him one of the most-read thinkers in the SaaS industry. Beneath the surface, nearly every post is structured using the McKinsey-born SCQA method: Situation, Complication, Question, Answer.

Several short sentences

His craft obsession goes down to the sentence level. Tunguz cites Verlyn Klinkenborg’s Several Short Sentences About Writing, which insists every sentence must earn its place and propel the reader forward. The rhythm of writing, like SaaS growth itself, matters.

Writing as a way to think about AI

Publishing daily also helps him think through frontier shifts like AI. Early on, readers told him they didn’t understand what an “agent” was. That feedback pushed him to create more tangible workflows and demos in his writing.

Inside companies, he notes, leaders should do the same: demo hours and hackathons are often the best way to spread AI literacy across teams.

A distribution-first thesis

Tunguz’s core investing philosophy can be traced back to a Belgian professor who wrote on the blackboard: innovation = invention + distribution. Early in his career, Tunguz weighted invention more heavily.

Today, he argues the compounding power lies in distribution.

With AI reducing build costs and customer acquisition costs year over year, even a 5–10% CAC advantage can halve the capital needed to scale.

How AI changes GTM

AI itself is reshaping distribution. In 2023, Tunguz predicted that half of consumer search would shift to AI. It took 18 months to come true. The lesson, he says, is that each platform shift rewrites the playbook for discovery. Google did it with search; app stores did it again on mobile. Now, in the age of AI, discovery itself may happen through agents; the new interface where products get found, recommended, and sold

Pricing is a perennial

Few topics have obsessed SaaS leaders like pricing, and Tunguz has written extensively about it. But despite a decade of essays, he still calls it “a perennial.” “It’s like the stock market. It changes every day,” he says. From seats to active seats to usage credits, every model is an experiment shaped by macro shocks, buyer power, and competitive dynamics.

The inevitability of better AI margins

On AI margins, Tunguz is optimistic. Efficiency curves are steep. Microsoft, for example, reported a 90% increase in tokens per GPU hour within a year. With innovations like caching, local models, and Stanford’s “Flash Attention,” he believes profitability will improve steadily. The real challenge for founders is to capture distribution early, even if it is uneconomical.

Uneconomic growth as a valid bet

In fact, Tunguz argues that investing ahead of profitability can be rational. Customers rarely switch vendors for at least three years, so in new categories, over-investing to capture distribution early can make more sense than optimizing for short-term margins.

He points to social networks, Google Search, and open-source projects as examples where distribution came first and monetization followed later.

Monte Carlo’s radical daily revenue model

That willingness to rethink fundamentals extends to SaaS metrics. Tunguz highlights Monte Carlo’s bold decision to abandon ARR/MRR and adopt a daily revenue model.

Account executives now log into dashboards that track revenue like a stock ticker, aligning the entire organization around real-time performance.

“PMF is grease”

Perhaps the most vivid lesson comes from Looker. With 150 permutations of product-market fit across verticals, geographies, and segments, the company found that product-market fit is not a binary milestone, but a continuous process.

In the “Year of Looker 500,” the team focused on extending its product-market fit to larger, 500-employee-plus enterprises, adjusting the product, sales motion, and implementation model to meet enterprise needs.

“PMF is grease,” Tunguz says. “It reduces the friction of scaling.”

That discipline powered 26 consecutive quarters of beating plan.

Why there’s never been a better time to start SaaS

All of this leads to a bold conclusion: Today may be the best time in history to start a SaaS business. AI is reinventing $1.5–$2T worth of workflows. Incumbents are unusually vulnerable. And buyers are eager to be educated. “The growth trajectories now possible were unimaginable five or ten years ago,” Tunguz reflects.

Umberto Eco and explaining China

The conversation closes, fittingly, with a discussion of books. Tunguz cites Umberto Eco’s “Chronicles of a Liquid Society,” a witty foil to tech’s optimism, and Dan Wang’s “Breakneck,” which frames China as a “nation of engineers” versus America as a “nation of lawyers.” Both reflect Tunguz’s conviction that the best founders and investors remain lifelong students.

Listen to the full conversation

Even more insights

Discover lessons from the field on using pricing to drive adoption, build sticky habits, and create an unfair advantage in AI markets.

Get the Pricing for AI Distribution Playbook →