Anatomy of a 250x: How Index Ventures' $3.5M Wiz Bet Became a $32B Google Exit, and What the Perplexity Markup Says Next

On 14 June 2026, the day Alphabet's $32B agreement to acquire Wiz became the headline number for every venture capitalist in the Western world, an AI search startup that did not exist four years ago was, by the same news cycle, raising money at $18B post-money. Perplexity is in early talks to raise $500M–$1B at roughly twice its December valuation, having reached nearly $100M in ARR and 15M+ active users. SoftBank Vision Fund 2, which reportedly wrote a $10–20M cheque into a $250M Perplexity round last year, is sitting on a paper return of around 5x. Index Ventures, which seeded Wiz with $3.5M in January 2020, is sitting on a paper return of 250x — $875M off one check, $4.3B off a $240M total position for a 13% stake in the company Google is buying.

Both stories, on the same week, are the same story. The Wiz deal is what a clean exit looks like in this cycle; the Perplexity round is what the next one will look like before the exit arrives. Read together they sketch the geometry of how capital is concentrating in a small number of conviction-led bets on infrastructure-shaped companies, and why the rest of the venture market — particularly direct-to-consumer brands still trying to compound through Meta — is being priced out of the trade.

The seed that became a $32B anchor

The Wiz story is not, strictly, a tech story. It is a timing story wrapped in a talent story wrapped in a regulatory story.

Sean Maguire, a Sequoia partner who worked on the firm's own cloud thesis before Index's Wiz bet landed, laid out the timeline in unusually granular terms. In 2015, cloud penetration of the Fortune 500 sat at 5–7%. By 2020 it was a first-class citizen inside the enterprise, and the security stack that enterprises had bought for on-premise workloads had become a stranded-cost liability. Wiz's wedge was asset discovery from outside-in — a thin layer of code that could enumerate every cloud workload a company was running without privileged access — and the larger product (cloud security posture management) was discovered in the first months of the company. The founding team's day-one solution was zero-trust networking; the day-180 solution was what Google bought. Maguire's framing of the trade is unusually candid for a partner still sitting on a similar bet: it was a market-plus-team bet, where the timing window did most of the work.

The team dimension is harder to replicate, and Maguire is blunt about why. Three of Wiz's four co-founders came out of Talpiot, the Israeli Defence Forces' elite unit that filters a small annual cohort through several years of combined military service, university coursework and engineering rotation. The argument, put plainly, is that Talpiot produces a concentration of cyber-talent that US universities — Maguire named Stanford and Harvard, both — no longer reliably produce, because admission standards and grade inflation have eroded the gap between elite and merely-credentialed candidates. Wiz could hire 20 ultra-elite engineers in its first quarter because the founders had spent a decade inside a pipeline that already filtered for them.

That combination — a market window, a small product that bought the right to discover the actual product, and a founding team with a captive recruiting funnel — is what Index bought for $3.5M in January 2020. Google has now agreed to pay $32B for the result, pending regulatory approval. The Index position becomes $4.3B; the firm turns a single seed check into 250x. It is the largest software M&A transaction on record, and the largest venture outcome Index has ever produced.

Perplexity and the new exit ceiling

If Wiz is the exit, Perplexity is the position being priced for the next one. The company is reportedly negotiating a $500M–$1B round at an $18B post-money, which would roughly double its December 2025 valuation. ARR is described as approaching $100M and active users as 15M+. Nvidia and SoftBank are inside the cap table.

The relevant comparison is not OpenAI at its last private mark; it is what the next two years of secondary trading and tender offers will reveal about the bid-ask between private marks and the public-market clearing price for AI applications. The SoftBank position is the cleanest data point. A $10–20M cheque written into a $250M round, marked at an $18B post-money, is a 5x+ paper return inside twelve months. The Perplexity round is reportedly doubling the December valuation in three months. At that compounding rate, the next round is a $30B conversation by year-end, and the round after that runs into territory where the only comparable exit is Alphabet buying Wiz or Microsoft buying OpenAI-adjacent infrastructure.

The question this raises is the one Sam Altman has now publicly answered. On a Stratechery interview relayed through the same news cycle, Altman was asked what would be more valuable in five years: a state-of-the-art model, or a 1-billion-daily-active-user destination site that doesn't have to pay for customer acquisition. His answer — that the destination site wins — is the most consequential narrative shift any frontier-model CEO has made on the question of where value accrues in the AI stack. If the OpenAI CEO believes application-layer distribution beats model-layer capability on a five-year view, then the Perplexity markup is not exuberant; it is the implied private-market price of the application thesis that Altman's own firm is no longer disputing.

The corollary, also relayed through the same discussion, is that bootstrapping a new billion-user destination site is now materially harder than it was in 2015. The structural advantages sit with incumbents that already own distribution — X, WhatsApp, Instagram, Google.com — and the private-market bid for any startup claiming it can become a destination is being priced off the assumption that it will have to buy its way in. The exit that justifies the price is, increasingly, an acquisition by one of those incumbents. Which is what Wiz was.

The other side: a D2C bloodbath and a privacy delisting

The Wiz exit and the Perplexity markup are what capital concentration looks like at the top of the cycle. Lower down, the same week produced a useful counterweight.

Connor and Sean Frank of Ridge Wallet, appearing on the same broadcast cycle, described the Meta advertising platform as a "bloodbath" for B2C brands. Three structural changes are doing the damage simultaneously. The default attribution window has been compressed from 28 days to 7, which destroys the signal brands used to use to allocate spend. Exclusion limits have been imposed, which means brands cannot easily suppress converters to find new prospects. And the algorithm is serving ads preferentially to warmer audiences — the existing customer base — which raises reported performance metrics while quietly starving the top of funnel. The iOS 14 changes, which stripped advertisers of cross-app signal, are still working through the system. The D2C consumer-confidence index hit its lowest reading since October 2024 on 18 March 2026.

The case study Frank offered is Solo Brands, the publicly listed parent of Chubbies and Solo Stove. The company is at a $12M market capitalisation on roughly $450M of revenue and more than $250M of debt. Chubbies, the surviving asset, is doing around $100M of revenue and is EBITDA-positive. Solo Stove is shrinking at roughly 25% year-over-year. The rollup was assembled by Bertram and Summit, both private-equity firms, and the operating record is a familiar one: the deal paid full price for peak multiples, the synergies never materialised, and the balance sheet now requires an asset sale to pay down debt. Chubbies is the asset most likely to be sold.

Ridge itself is the case study of what a brand that read the platform shift correctly looks like. The company has spent more than $250M on Meta ads cumulatively, per Frank, and 15% of its orders now run through Affirm. Frank is also blunt about the power asymmetry: after a quarter-billion in Meta spend, he has never personally received anything from Mark Zuckerberg, and gets more executive attention from Evan Spiegel at Snap. The operational takeaway is that D2C at scale now requires a media-relations strategy aimed at the platforms themselves, not just a buying strategy aimed at their ad inventory.

Outside the venture cycle entirely, the US Treasury's decision to delist Tornado Cash from the OFAC sanctions list was the other structurally significant item in the week. Coin Center and Brian Armstrong sued the government over the original 2022 designation, and the delisting was framed by crypto-libertarian commentators as a privacy win. The delisting does not affect the criminal prosecution of Tornado Cash co-founder Roman Storm, whose trial is scheduled for April on charges of operating an unlicensed money-transmitting business, conspiracy to commit money laundering and sanctions evasion — counts that carry up to 45 years. Storm has called the prosecution a "terrifying criminalisation of privacy." The delisting resolves a regulatory overreach question; the trial resolves a criminal-liability question, and the two are not the same.

The structural read

The shape of the cycle is now legible. Capital is concentrating into a small number of conviction bets at the infrastructure layer — cloud security, frontier-model applications, AI-native search — and exiting through strategic acquirers that already own distribution. The Wiz deal is the template: a market window, a Talpiot-credentialed team, a wedge product that bought the right to discover the real product, and a buyer whose existing distribution makes the asset instantly accretive. The Perplexity markup is the next position being priced on the assumption that the same template applies. Index Ventures is the firm that has now demonstrated the template twice in the same decade.

What the rest of the venture market does with that information is the open question. The D2C brands still trying to compound through Meta are working with a deteriorating ad platform, a tougher attribution regime, and a consumer-confidence index at its weakest reading in six months. The PE-backed rollups that paid peak multiples for consumer brands are underwater on their debt and are now in the asset-sale phase of the cycle. The mid-cap SaaS companies that were priced for 2021 growth rates are being quietly marked down in secondary transactions that do not yet show up in the headlines.

The trades that are working — Wiz, Perplexity, the AI-application tier more broadly — share three characteristics. They are infrastructure-shaped rather than brand-shaped, meaning they win or lose on technical merit rather than on customer-acquisition cost. They are concentrated rather than distributed, meaning a small number of high-conviction bets rather than a portfolio of correlated ones. And they have an identifiable strategic acquirer whose distribution makes the exit inevitable rather than optional. The capital is voting for a specific kind of company, and the rest of the market is being repriced whether it likes it or not.

The Celtics sale, in the same week, for $6.1B to a buyer with no public profile and a minority owner publicly questioning whether the buyer has the money, is a small reminder that even record transactions can carry financing structures that will be tested in the next downturn. But that is a different cycle. The cycle that is currently being priced is the one Index just cashed a 250x ticket on, and the next ticket is being written at $18B post-money with a 5x return for the lead investor inside a year. The pattern is no longer rare. It is the new ceiling.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://www.youtube.com/watch?v=yT3K7hMRA1c