Going for broke: how Menlo's Venky Ganesan plays a market that 'is not venture anymore'
Key points
Key takeaways from the 20VC interview with Menlo Ventures partner Venky Ganesan (October 2026):
The lesson from a 90% loss. Two decades ago, early in his career as a venture associate, Ganesan put $5,000 into Avanex (AVNX) at its IPO; the position climbed to roughly $200,000 before collapsing 90%, and he sold for around $8,000-9,000. His takeaway: taking chips off the table makes sense when a windfall is life-changing, but the advice depends on your balance sheet — and at Menlo today the firm is deliberately going for broke.
Going for the grand slam. At this point in Menlo's history the firm is playing to win everything rather than protect anything: it wants to be inside the defining AI companies of the era, it has already banked home runs, and it is prepared to take concentrated, swing-for-the-fences risk with fund capital.
Venture is not what it was. Ganesan's team came back reporting they could not find anything under $100 million — that being the size of seed rounds, not valuations — while some new AI labs want to raise billions. He acknowledges the moment is disorienting, but argues a dot is not a line: long-term strategy should never be drawn from a snapshot of a market that changes this quickly.
Dance while the music plays, but differently. Invoking the Chuck Prince quote, Ganesan warns that venture firms which stepped out in 1996-97 missed 1997-99 and climbed back in at the top in 2000. The answer is to keep playing but more selectively, using portfolio composition and position sizing to cushion the turn of the cycle.
Seed as an options book. Each seed investment is an option bet bought to discover an outlier, so funds need enough at-bats; only when revenue and quantifiable metrics prove the outlier should a fund size up its position. Large funds are increasingly indifferent to seed valuations because they are buying a cheap seat at the table, with the real goal of scaling in later.
Metrics get gamed, every cycle. Any metric investors weight heavily will be gamed — Ganesan cites net revenue retention, where a $10 purchase order followed by a $50 PO a week later flatters the ratio versus a single $100 PO. In a boom the "bezel" is high, so the tell is whether founders build terminal value or just chase markups.
Kingmaking is Soros reflexivity. Fast markups attract more capital, notoriety and talent, which compound further markups — a virtuous loop until it stops, as all reflexivity eventually does. Cycles usually crack on a major debt default rather than equity writedowns, because the players who break are the leveraged ones: at 4x leverage you must be right on timing, not just direction.
Leave the ego at the door. Ganesan admits ego has cost him — walking away from small pieces or haggling over valuation — and says the only ego that matters is returning money to institutional LPs. If a founder can take Peter Fenton's money at half the price, so be it; the job is making money for investors, and the rest is noise.
Read more: 20VC
