Instacart Series I

Andreessen Horowitz and Sequoia betting $265M at a $39B valuation on grocery delivery in March 2021—literally peak "we're never leaving our houses again" energy—represents venture capital's most exquisite demonstration of recency bias packaged as strategic foresight.[1] The fundamentals weren't *bad* per se: Instacart had genuinely scaled to profitability-adjacent territory (read: positive unit economics if you squint at the EBITDA and ignore all the basket subsidies), commanded real market share in a sector that historically made investors weep into their term sheets, and had transformed from "Uber but for groceries lol" into actual infrastructure that Kroger and Safeway couldn't ignore. But $39B? That's 7.5x the $5.2B valuation from just eighteen months prior, a markup that assumes pandemic shopping behaviors would calcify into permanent human psychology rather than, you know, revert to mean the second Pfizer started jabbing arms.[2] The growth metrics were legitimately impressive—GMV probably north of $15B annually, take rates hovering around 10-12%—but you could feel the desperation in the cap table expansion, like they knew the music would stop eventually.

[1] To be clear: I have nothing against founders capitalizing on temporary insanity. That's literally the game. But let's not pretend this was 4D chess when it was really just "holy shit people will pay us to shop for them AND investors will pay us to let them pay us."

[2] The comparable here is obvious and painful: DoorDash IPO'd at $39B in December 2020, Instacart looked around and said "yeah, we're basically that," which—fine, arguable—but DoorDash's restaurant delivery model had stickier network effects and better margin expansion potential than the soul-crushing economics of frozen peas and twelve-packs of Charmin.

The investor signaling here oscillates between "legitimately impressive" and "ominous harbinger of froth"—and yes, I'm aware that sentence could describe literally any late-stage deal from 2020-2021, but bear with me because the specifics matter.[3] Andreessen and Sequoia co-leading means you've got Silicon Valley's equivalent of The Beatles reuniting for one more stadium tour: brand names that *actually* move markets, networks that *actually* help with IPO roadshows, and pattern-matching that *actually* works until it catastrophically doesn't. The party round nature—fifteen total investors crammed into the cap table like clowns in a Tesla—suggests either extraordinary conviction or extraordinary FOMO, and I'll let you guess which. What saves this from full-on bubble territory is that Instacart wasn't some pandemic-native upstart speedrunning Series A to I in eleven months; they'd been grinding since 2012, had real competitive moats (proprietary shopper networks, retailer integrations that took years to build), and had successfully fought off the Uber/Postmates/DoorDash incursions into their territory. The deal wasn't *stupid*—it was just optimally timed for maximum seller advantage and minimum buyer price discovery.

[3] I spent three hours trying to find comparable SaaS multiples to justify the valuation and eventually gave up because Instacart isn't really SaaS (though they'd love you to think their "enterprise" retailer tools are), isn't really marketplace (though the economics rhyme), and isn't really logistics (though the actual work is). It's this beautiful Frankenstein's monster of business models that defies clean Excel modeling, which is either brilliant or convenient depending on your position in the cap table.

The 7.1 rating exists in that uncomfortable space between "objectively solid execution" and "probably fucked on exit pricing"—a purgatory I'm increasingly convinced describes 60% of 2021 vintage venture deals.[4] Instacart *did* subsequently file for IPO at a $10B valuation in September 2023 (!!!) before eventually going public at $9.9B that same month, which means anyone who bought into this Series I ate a 75% markdown in thirty months, a loss pattern that would get a mutual fund manager fired but somehow counts as "still a win because we got liquidity" in VC-land. Yet—and here's where the 7.1 justifies itself against pure cynicism—the company survived, shipped, and exited while Gopuff imploded, Getir retreated from the U.S., and countless other delivery plays simply vaporized capital into the quantum foam. The business model *worked*, just not at $39B valuations. The fundamentals were defensible, the execution was professional, and the investors weren't idiots—they were just playing a game where being 18 months early and 75% wrong on price still beats being right but uninvested.

[4] Look, I'm not saying everyone who deployed capital in 2021 should be banned from managing money, but maybe we could implement a brief cooling-off period? Like how casinos cut you off after you've been at the blackjack table for fourteen hours straight? Just spitballing here.