Cem Sertoglu on UiPath, Venture Risk and Backing Founders

Harry Stebbings asks Cem Sertoglu how he became a venture investor. Sertoglu founded social-networking software company SelectMinds in New York in 1999, survived the technology crash and eventually exited. After moving back to Istanbul, where he grew up, he began meeting and investing in young Turkish consumer internet companies.

He believed consumers in different markets would ultimately behave similarly online, even where others doubted it. Several early angel investments became major Turkish exits, including an e-commerce business sold to eBay and a food-delivery company sold to Delivery Hero. He estimates his overall angel multiple at around 10 times, but says he has not calculated it.

Sertoglu raised a fund because he could syndicate the first few million dollars but struggled to find regional investors for rounds above $5 million to $10 million. He wanted to support companies in Turkey and Eastern Europe through subsequent rounds, fearing some otherwise might fail to find financing.

Capital from outside the region tends to rush in and out with the cycle, he says. Local familiarity can help an investor assess a founder’s background and references rather than rely only on a polished pitch. Even when prices rise, Sertoglu believes a ten-year fund should keep investing in strong opportunities instead of attempting to time the market.

Stebbings raises Doug Leone’s argument that venture has become a commoditized, low-margin industry. Sertoglu disagrees: cash may be interchangeable, but the relationship and support a founder receives differ. He distinguishes early-stage partnerships from the asset-management businesses some multi-stage firms have built alongside them.

Stebbings counters that founders often prefer faster cash, higher prices and less involvement, making large multi-stage firms formidable competitors. Sertoglu says some founders want exactly that, while others choose partners almost as they would co-founders. His preference for a small partnership reflects the business model of his own firm.

He will compete for a sought-after founder when there is mutual chemistry and an understanding of what his firm offers. Looking at exceptional outcomes, he says paying twice the initial valuation would still have produced an attractive result. UiPath is his example: he reports $16.5 million invested across rounds and $2.1 billion in proceeds, not merely a paper valuation.

He says fund one has a 20-times multiple on invested capital, with 2.7 times attributable to investments other than UiPath and about 85% of fund proceeds from UiPath. The figures are his reported fund metrics; the interview does not provide an independent audit or a complete breakdown of remaining assets.

Despite saying price matters less in extreme winners, Sertoglu says the firm must still assess whether an entry price is fair. It has passed on price before, with mixed results. He cites missing Bolt, then called Taxify: the team admired the founder but could not see the small Baltic ride-hailing company becoming a global contender and questioned the market’s take rates.

Stebbings describes passing on companies because he disliked the business despite admiring their founders. Sertoglu ranks founder first, market second and traction a distant third at his stage. Founder character and values matter most; the market must still be large enough for an investment to return the fund if it succeeds.

The firm models potential outcomes, but Sertoglu says early investment memos can look comically conservative after an exit. UiPath had seemed capable of returning its fund; he says it ultimately returned about twelve times the fund. Scenario planning provides discipline about market size, not reliable precision.

Sertoglu says each partner typically makes four or five investments per fund, allowing time and attention for each founder. The firm often seeks a board seat if its check and ownership warrant one. Its first check typically buys 10–20%, but it can accept less or more if the economics make sense.

Fund one made 15 investments and fund two 18, mostly seed and Series A, with selected pre-seed and Series B checks. The small number means ownership and opportunity cost still matter: the team has passed on companies when the stake was too small for its limited number of investments.

Sertoglu believes investor signaling is real. When assessing a company, his team looks at the cap table and the behavior of existing shareholders, including whether they follow on. He also sees premature assumptions of product-market fit as a frequent reason companies fail to scale efficiently.

Startups can mistake fundraising and headcount growth for validation, then build a sales team around a script that is not yet right. Once the company scales that playbook, it can become harder to adapt and iterate. He argues growth should follow evidence that the product and go-to-market approach work, rather than precede it.

His firm practices being quiet on boards: members should prepare obsessively and understand the business, but choose carefully when to express a strong opinion. He warns that the pressure to demonstrate value can lead investors to repeat aphorisms or speak too often, even though every company follows its own trajectory.

For sensitive issues, Sertoglu prefers a relationship that allows a private conversation with the founder before a matter appears for the first time at a board meeting. Once the important issues are selected carefully, he believes they can be raised with the full board. Founders usually know the business better and have more at stake, so investors should support them when needed.

Stebbings argues that cash is a commodity and venture firms must market themselves. Sertoglu replies that his firm sells time, attention and capacity over a long partnership. He says the best founders benefit from smart people aligned behind them, although he hopes rather than claims every board is valuable.

On early valuations, he agrees that raising at a price a founder can plausibly increase two to three times in the next round can be a useful rule of thumb, depending on stage. He regards early-stage financing less as a short-term trade than as a long-term alignment between a founder and investors.

His team tries to encounter at least one disagreement during term-sheet discussions to understand the relationship dynamic. Protracted or contentious negotiations over governance can leave distrust on both sides. The firm will pay high prices, he says, but refuses to be rushed when it cannot understand a business; it likely misses some good companies as a result.

Asked whether expensive 2021 checks produced high-quality companies, Sertoglu says it is too early to judge the whole portfolio: some have grown into their valuations and others are still developing. Stebbings shares his own experience of some well-funded seed companies losing focus and struggling to raise the next round.

Round labels have become inconsistent, he says. He calls a company with little beyond an idea pre-seed, regardless of the size of the check, and would invest if convinced by founder and market. High-risk investing should feel uncomfortable, he argues; even after years of experience the work has not become easy.

Stebbings asks whether wealth makes investors more willing to take risk. Sertoglu says it can help, but an LP challenged the firm’s general-partner commitment of nearly 10% to a $250 million fund: too much personal exposure could make managers excessively risk-averse. He accepts the concern and says a commitment must be judged relative to an investor’s means.

He says his own ability to make a large commitment depended on the first fund’s outcome, so it would be unfair to impose the same hurdle on a young emerging manager. His LP base is diverse, he says, but not dominated by large North American institutions.

Sertoglu argues venture should not be classified geographically in the same way as private equity: a company may originate in Eastern Europe or Turkey but seek global customers and exits. He says his own portfolio has not suffered the political or currency risks some LPs fear. That is his experience, not proof that such risks cannot affect other companies.

He says the firm has lost deals on price when another investor accepted a lower expected return, and on desired ownership when existing investors would not allow a new lead enough of the cap table. Even a firm that emphasizes relationships must still underwrite opportunities to its own return requirements.

Sertoglu met UiPath founder Daniel Dines in Bucharest in 2014, when the company had about 12 people and had grown from a consulting and automation business. Dines impressed him through technical depth, customer understanding and practical attention to the next feature, while also seeing broad potential applications.

The team spent six to nine months getting to know Dines before investing in 2015. They initially worried whether he could lead go-to-market, citing his less polished English and storytelling at the time. Through interactions with potential senior hires, they came to believe he could lead the company without resolving that question through a separate co-founder.

The initial plan was to raise $1.5 million. Sertoglu says his firm led with $1 million, Credo added $500,000 and Seedcamp another $100,000, for about $1.6 million at just under a $7 million pre-money valuation. The company had around $500,000 in revenue, after years in an earlier, more services-oriented business.

Although the fund could have written the whole round, it wanted partners and reserves for a company that might struggle to attract later investors. A year or so later, he says, roughly 40 firms passed on UiPath. Existing investors bridged the company because they could see customers getting value, even though the headline numbers were not yet exceptional.

Dines and Sertoglu remember a disagreement over pricing a round: he says the market’s poor response led them to use a convertible rather than agree on a priced round. The $2.5 million bridge eventually converted at around a $60 million valuation, he recalls; with hindsight, he would have preferred to double down in a priced round sooner.

When traction became visible, outside interest reversed. Accel led the Series A at an approximately $80 million valuation, and Sertoglu’s firm invested another $3 million. He says Accel’s growth team later led a round at a $1 billion valuation, where his firm wrote a further $10 million check after a difficult internal decision.

Japan’s rapid adoption helped build conviction that other markets could also scale. Sertoglu believes UiPath grew from $1 million to $100 million in annual recurring revenue in 21 months, which he recalls as exceptionally fast at the time. That is his recollection of ARR growth, not a verified measurement of recognized revenue.

He distinguishes the overlooked-company bridge from the later decision to follow momentum at a $1 billion valuation. The firm skipped the next round at $3 billion, led by Sequoia, and began selling some shares around a later $7 billion valuation to realize gains for LPs.

He recalls selling small portions of the holding in several transactions, generally less than 10% of the stake each time, at premiums to the primary rounds. Ownership peaked around 18% and was slightly below 10% at IPO, reflecting both sales and dilution; he cannot reconstruct the exact split. He says the firm had already realized around three to four times the fund by then.

After the IPO, the firm sold more quickly and distributed shares in kind to LPs who preferred stock. Sertoglu argues that an early-stage investor’s informational edge decays once a company is public and that fund lifetimes also constrain holding periods. This is his approach, not a universal rule for managing public positions.

Asked whether a huge success makes a team less motivated to keep working, he says people want to build a firm they can eventually inherit. He wants an institution that outlasts him. He also thinks a ten-year life is often too short for a seed-focused fund, though the deadline creates useful pressure to consider liquidity.

In hindsight he wishes the firm had participated in UiPath’s $3 billion round and might have used a different post-IPO divestment strategy. The outcome remains exceptional, but he separates that result from the decisions that could have been better.

Sertoglu says only three of 15 fund-one investments had failed to return capital, while six companies remained active and were expected to make money. The latter is a forecast, not realized proceeds. He wonders whether some lower-risk, capped-return investments consumed resources that could have backed higher-risk opportunities.

Although he still thinks about downside protection, he considers it overrated for early-stage investments. Liquidation preferences and similar terms can be renegotiated at exit, he says, and have not materially changed the outcomes he has seen. This is his experience rather than a claim that investor protections never matter.

His largest realized loss was about $6 million in an HR business whose plans became difficult after regulatory changes. He thinks the firm should have scrutinized a follow-on check more carefully; the lesson was to evaluate reserves and additional commitments rather than assume an existing investment deserves more money.

On weak IPO and acquisition markets, he would worry if the trend persisted, but expects cycles to turn. He says fund vintage strongly affects performance without excusing managers from explaining their decisions. He suspects many 2021 funds may fail to return capital, while acknowledging he has not examined the underlying data.

Stebbings worries AI pricing resembles or exceeds the 2021 bubble; Sertoglu agrees prices are concerning but is unsure conditions are uniquely difficult. His firm’s recent investments all use AI in some way, he says, but it is not participating in huge foundation-model rounds. His comments are market judgments, not a prediction of individual company outcomes.

In the quick-fire round, he repeats that early-stage financing is a long-term alignment rather than simply a trade. He credits Fred Wilson as an early source of encouragement and learning, and admires USV and Benchmark for maintaining discipline about their strategy.

He urges founders to recognize their leverage on the cap table rather than over-negotiate remote scenarios. In deciding whether dilution is worthwhile, he suggests asking whether the investor meaningfully improves the probability of reaching the company’s vision. This is a heuristic, not a precise way to calculate an investor’s contribution.

Sertoglu worries technology has concentrated resources in too few hands and expects the problem to worsen before improving, though he hopes compatible solutions will emerge. His final lesson is patience: venture outcomes take a long time. He wishes more founders asked investors for data on their follow-on behavior as a signal of support.