3i Member Spotlight

The episode contains genuinely useful VC mechanics (fund math, exit timing calculus, the agentic AI primer) but roughly a third of runtime is consumed by a religious/philosophical tangent and personal anecdotes that deliver zero actionable value to a B2B operator. The Double Verify hold-vs-sell analysis is the densest segment.

“say a fund is about 225 million...we have to have six companies that are worth a billion dollars each. Because then our yield would be a hundred million times six is the 600 we need to get to the 3x net to our LPs”

“McKinsey estimates that 57% of work hours will be transformed from human labor to agentic software”

The claim that investor proximity is inversely correlated with startup success is a genuinely counterintuitive finding, and the 'hold the dragon' framing of the Double Verify decision is vivid. Otherwise the episode recycles standard VC wisdom, Wayne Gretzky puck clichés, and AI talking points without adding first-principles analysis.

“the success of the startup was inversely correlated to the proximity of the investors to the entrepreneur”

“when you find these dragons, I say, like, hold on to the tail, let it fly some more”

Both guests are genuine multi-decade practitioners with auditable outcomes - first investors in Nutanix, DoubleVerify, and Braze from pre-revenue to IPO - giving them real credibility as operators, not thought-leaders. They are not top-decile name-recognition VCs, and the conversation exposes some limits in depth, but the track record is real.

“companies like Nutanix, Braze and Double Verify. These are all companies that went from pre product, pre revenue, we were the first investor to going public”

“we had 8 million invested. It was worth about 64 million. We sold off 24 million and we rolled 40 into the new, um, equity group...a total return of I think 72x”

The Double Verify exit story is unusually concrete - named company, dollar amounts, ownership stakes, round-by-round logic, and final return multiple - and the Engle Health example adds a named, sourced company with revenue and efficiency figures. The AI section leans on third-party McKinsey data and loses some points for unverified stats, but the debt-collection bot anecdote (12-15 vs. 126 resolutions per week) is a sharp, specific proof point.

“the average American employer spends $170,000 a year per employee...the market capitalization of those companies that make that software is about $16 trillion”

“approaching $1 billion in top line revenue. They are already profitable and they're doing it 5x more efficiently than other businesses”

The host occasionally pushes back (challenging the exit-discussion question, interjecting on patents) and asks a sharp structural question about hold-vs-sell, but he repeatedly completes guests' sentences, lets a 20-minute religious tangent displace the stated AI agenda, and closes with an admission that the episode's headline topic was never reached - a significant craft failure.

“I'll challenge you on that”

“It's been an hour and what a great conversation. I'm sorry we didn't get to AI”