Ecommerce · Deep dive

Instacart

The grocery marketplace that survived Amazon, rode the pandemic to a $39B private valuation, IPO'd at $10B — and now defends a shrinking delivery share against DoorDash with an ads engine, smart carts, and enterprise software.

at risk

The headline numbers still grow, but the growth is increasingly bought — $10 minimum baskets, waived fees, flat order values — while DoorDash has already passed Instacart in third-party grocery-and-retail order volume, retailers multi-home, and the stock sits where it IPO'd three years ago.

My take ★★ Very interesting ★ Interesting − Maybe ✗ Not interesting

Snapshot

Instacart — legally Maplebear Inc. — is North America’s largest dedicated online grocery marketplace, connecting roughly 1,800 retail banners and over 100,000 stores with about 600,000 gig workers who pick and deliver orders. In FY2025 it processed $37.2 billion of gross transaction value (GTV) across 338.8 million orders for $3.74 billion of revenue, over $1 billion of which was high-margin advertising sold to CPG brands. Q1 2026 was its first-ever quarter above $10 billion GTV and $1 billion revenue (May 2026). Yet the market values the company at roughly $10 billion in July 2026 — almost exactly its September 2023 IPO valuation and about 75% below its March 2021 private peak of $39 billion. The stock price is asking whether Instacart is a compounding ads-and-infrastructure platform or a pandemic-era intermediary squeezed between DoorDash’s scale and its own retailers’ urge to disintermediate it.

Founding story

Apoorva Mehta is the patron saint of persistence-over-insight founding stories. Born in India, raised partly in Libya and Canada, trained as an engineer at Waterloo, he worked as a supply-chain engineer at Amazon in Seattle before quitting in 2010 to start companies. He then failed roughly 20 times — an ad network for games, a Groupon-for-food, a social network for lawyers — before, in summer 2012, staring into a San Francisco fridge containing only a bottle of sriracha. Groceries were the one major retail category ecommerce hadn’t cracked. He coded the first version in about three weeks, missed Y Combinator’s summer 2012 deadline, and got in anyway by using the app to deliver a six-pack of beer to YC’s partners; he met co-founders Max Mullen and Brandon Leonardo around the batch and raised $2.3 million within months.

The founding insight was contrarian timing: Webvan incinerated ~$800 million proving warehouse-based grocery delivery didn’t work in 1999, but by 2012 smartphones let Instacart use existing supermarkets as warehouses and gig workers as labor — near-zero capex. Mehta ran the company nine years, handed the CEO role to ex-Facebook executive Fidji Simo in August 2021, and exited the board at the 2023 IPO. Simo left for OpenAI in 2025, succeeded by insider Chris Rogers, an Apple veteran who had run retailer partnerships since 2019.

How it works

A customer opens the app, picks a store — Costco, Kroger, Publix, Aldi, Sprouts, Restaurant Depot — and builds a cart from a catalog Instacart syncs with the retailer’s inventory and pricing feeds. The order hits a dispatch system that offers it to nearby gig shoppers as a “batch,” often bundling two or three customers’ orders into one store run. The shopper walks the aisles with the app, scanning items, messaging the customer about substitutions, checking out with a company payment card, and driving the delivery. Instacart’s algorithms handle batching, routing, substitutions, and fraud; the retailer just keeps shelves stocked. The crucial physical fact: Instacart owns no inventory, warehouses, or vehicles — its assets are catalog data, demand aggregation, and the dispatch engine.

Around that core sits an enterprise stack. Storefront and Storefront Pro white-label the whole experience as the retailer’s own site and app (Restaurant Depot relaunched on it in October 2025). Carrot Ads lets retailers run their own retail-media businesses on Instacart’s ad tech. Eversight (acquired 2022) runs AI pricing and promotion experiments. Caper Carts — from the ~$350 million Caper AI acquisition in October 2021 — are camera-and-scale smart shopping carts that recognize items as they’re dropped in, bringing ads and personalization into physical aisles; a wave of independent grocers adopted them in late 2024. FoodStorm handles catering. The strategy: even if delivery commoditizes, Instacart becomes the software layer under grocery itself.

Product and business overview

Four named businesses. The Marketplace (instacart.com and app) is the consumer demand engine — same-day delivery and pickup from ~1,800 banners, plus a restaurants tab powered by an Uber Eats partnership expanded in May 2025. Instacart+ is the $99/year membership with free delivery, now on baskets as small as $10 (lowered from $35 in 2025). Instacart Ads sells sponsored-product auctions, display, and coupons to more than 7,000 CPG brands that pay to win placement in search results — the profit engine, at $1.065 billion in 2025. The Enterprise Platform (Storefront, Carrot Ads, Eversight, Caper, FoodStorm) sells the picks-and-shovels to retailers, including nominal competitors’ media networks. A February 2026 partnership with Toast extends the B2B push into restaurant supply, and Instacart Business serves office and foodservice buyers. In December 2025, Instacart became the first app with checkout embedded directly inside ChatGPT.

Business model and pricing

Revenue books in two lines. Transaction revenue — fees from retailers (roughly 5-15% of order value depending on the deal) plus consumer fees — ran at 7.1% of GTV in Q1 2026 ($733 million). Advertising and other revenue ran at 2.8% of GTV ($286 million), making the all-in take rate about 9.9%. Consumers pay a delivery fee (from $3.99, waived for Instacart+ members on $10+ baskets), a service fee typically 5-10% of subtotal, and — the quiet one — item markups: The Markup’s much-cited study found prices averaging 17.5% above shelf across 11 retailers, and consumer guides in 2026 estimate a full basket costs 40-50% more all-in than shopping in person. Instacart+ costs $99/year or $9.99/month. Shoppers are paid per batch — a floor of roughly $7-10 plus tips, and base pay covers the whole batch even when it contains three orders. The structural tension: every party pays — the consumer via markups and fees, the retailer via commission and ceded data, the brand via ads — which works while Instacart is the only aggregator that matters and gets renegotiated the moment it isn’t.

Traction over time

Read the mix, not the headline: orders have grown faster than GTV since 2024 because average order value is flat-to-down — $112 in Q4 2025 (-1%), dented by $10-minimum free delivery and small restaurant baskets. Volume is being bought with margin concessions. Meanwhile the company is genuinely profitable (net income of $457 million in 2024; adjusted EBITDA up 23% in 2025) and repurchased $1.4 billion of stock in 2025.

Market analysis

Brick Meets Click puts US eGrocery at roughly $127-130 billion for 2025 — about 14-15% of total US grocery — with a record 61% of households buying groceries online by mid-2025. Delivery is the fastest-growing fulfillment mode, up 30-36% year over year in mid-2025 months, and online grocery is projected to grow ~8.9% annually through 2029 to ~17% of grocery sales. Structural forces are favorable: habitual post-pandemic adoption, restaurant-to-grocery substitution, and retail-media budgets migrating to platforms with closed-loop attribution. The problem is that the growth no longer accrues mainly to Instacart: pickup (38% of eGrocery, dominated by Walmart and Kroger first-party) bypasses it entirely, and delivery growth is contested by DoorDash and Uber, for whom grocery is an incremental attach to an existing logistics network rather than a standalone P&L.

Competitive intel

The competitive story is one number: Wedbush estimated Instacart’s share of US grocery delivery fell from about 70% in 2023 to 58% in 2024, and by 2026 DoorDash claimed — citing YipitData — to be the top US third-party marketplace by order volume in grocery and retail. DoorDash (market cap several times Instacart’s) added Kroger’s ~2,700 stores in October 2025, holds 67% of restaurant delivery, and amortizes courier density across restaurants, convenience, and grocery. Uber Eats runs the same playbook off the rides base, with eMarketer forecasting its US grocery/retail GMV ($27.6B) roughly matching Instacart’s by 2025. Walmart keeps delivery in-house behind the $98/year Walmart+; Amazon dropped Instacart from Whole Foods after the 2017 acquisition and keeps pressing same-day Fresh; Target owns Shipt. Instacart’s genuine advantages: the deepest catalog integrations (real-time inventory, EBT/SNAP, pharmacy, loyalty-card sync), the big-basket weekly shop where its shopper base and substitution flows beat restaurant couriers, club-store partnerships (Costco, Sam’s Club, Restaurant Depot), and an enterprise/ads stack no courier matches. Its weakness: the consumer app is the profit pool, and that is exactly what DoorDash attacks with cheaper, bundled delivery.

History and evolution

• 2012 — Founded in San Francisco; YC S12 via the beer-delivery stunt; $2.3M raised.
• 2014 — Series C at ~$2B; expansion beyond the Bay Area accelerates.
• 2017 — Amazon buys Whole Foods, then drops Instacart (its partner since 2016, reportedly ~10% of revenue); terrified grocers flock to Instacart in response.
• 2019 — Tip-subsidy scandal: Instacart admits base pay had been offset against customer tips; Mehta apologizes and reverses it.
• Mar 2020 — COVID demand up ~300%; shoppers stage a nationwide strike over hazard pay and PPE; company hires hundreds of thousands.
• Jan 2021 — Lays off ~1,900 in-store employees, including its only unionized shoppers.
• Mar 2021 — $265M round at $39B; Fidji Simo becomes CEO in August; IPO plans shelved as growth stalls.
• 2021-2022 — Buys Caper AI (~$350M) and Eversight; marks internal valuation down four times, ending 2022 near $10B.
• Sept 19, 2023 — IPO at $30/share (~$10B); opens at $42, closes day one at $33.70; PepsiCo invests $175M.
• 2024 — First full GAAP-profitable year ($457M net income); grocery-delivery share slides to ~58% (Wedbush).
• May-Aug 2025 — Simo leaves for OpenAI; CBO Chris Rogers becomes CEO Aug 15; Uber Eats restaurant partnership expands; $10 minimum baskets for Instacart+ free delivery.
• Oct 2025 — Restaurant Depot relaunches on Storefront Pro; DoorDash-Kroger deal lands the same month.
• Dec 2025 — First app with native checkout inside ChatGPT via the Agentic Commerce Protocol.
• Feb-Mar 2026 — Consumer Reports/Groundwork investigation exposes AI price-testing that charged different customers different prices; Instacart ends the tests. OpenAI kills Instant Checkout in March, refocusing agents on discovery with checkout back in partner apps.
• May 2026 — First-ever $10B GTV / $1B revenue quarter reported.

What people say

The case for. Bulls — including value managers like Miller Value Funds (2025) — point to nine straight quarters of double-digit GTV growth through Q1 2026, ads reaccelerating to 16% growth (fastest since Q3 2023), real GAAP profits, roughly $1.4 billion of 2025 buybacks against a ~$10B market cap, and an enterprise strategy that converts would-be disintermediation into software revenue: retailers that leave the marketplace often keep Carrot Ads or Storefront. Consumers praise selection depth and the substitution/chat flow for large weekly shops — the use case couriers handle worst. The ChatGPT integration made Instacart the default grocery rail for the largest AI assistant, and management argues its catalog data makes it the natural agent-commerce backend (PYMNTS, May 2026).

The complaints. Shoppers are the loudest: Facebook and Reddit communities document batch pay of $7-10 covering up to three orders, earnings that fell from $40-50/hour pandemic peaks to sub-minimum-wage batches, and opaque pay changes the company won’t explain (CBC, TechCrunch, 2021-2023; the 2019 tip-subsidy scandal remains the reference point). Customers cite markups averaging 17.5% (The Markup) and all-in costs 40-50% above shelf, and the February 2026 Consumer Reports dynamic-pricing exposé — 74% of tested items showing multiple simultaneous price tiers — forced a public retreat. Grocers, per Mercatus and NielsenIQ commentary, resent ceding customer data and vendor-funding dollars to a middleman while running 2% margins. The bear thesis (Seeking Alpha sell notes, 2025-2026; Wedbush share data) is simple: DoorDash’s bundled logistics wins the marginal order, AOV stagnation shows Instacart buying volume with concessions, and regulatory pay floors in Seattle and NYC keep raising costs.

Outlook: well positioned or at risk?

At risk. The composition of Instacart’s growth betrays it. Orders grew 15% in 2025 against 11% GTV growth; Q1 2026 order value rose only because club-store and Restaurant Depot mix propped it up. Free delivery on $10 baskets is a defensive subsidy dressed as a product feature — it exists because DoorDash and Uber bundle grocery into memberships consumers already have for restaurants, and Instacart has no adjacent demand pool to bundle from. The share math is already public: ~70% of US grocery delivery in 2023, ~58% in 2024, and DoorDash claiming the #1 spot in third-party grocery-and-retail order volume by 2026 while signing Kroger. Meanwhile the highest-margin swing factor, ads at 2.8% of GTV, depends on the marketplace keeping the eyeballs the couriers are taking, and the enterprise business — genuinely good — monetizes retailers’ desire to need Instacart’s marketplace less. The counter-case is real: profitable, cash-generative, buying back nearly 15% of its market cap a year, with the best catalog data in grocery and a plausible seat in agentic commerce. But equity sitting at its IPO price after three years of double-digit growth means the market already suspects a position being harvested rather than compounded, and the evidence — basket stagnation, share loss, a CEO who left for the platform that may intermediate her old company — says the market is right. Well-run, profitable, and structurally losing the war for the marginal grocery order.

How a challenger would attack it

Attack the markup, not the delivery. Instacart’s profit pool rests on every party paying — 17.5% average item markups, service fees, $99 memberships, retailer commissions of 5-15%, and an ads auction — which means a full basket runs 40-50% above shelf price. That umbrella is the target. A challenger doesn’t need to out-logistics DoorDash; it sells transparent shelf pricing as the product, either by partnering with grocers who resent ceding data and vendor dollars at 2% margins, or by white-labeling first-party fulfillment the way Storefront does but without the marketplace tax attached. The trust wound is fresh: the February 2026 Consumer Reports exposé found 74% of tested items carrying multiple simultaneous AI-tested price tiers, forcing a public retreat — a challenger runs “one price, the shelf price” against that headline forever. The second front is labor: shoppers earning $7-10 batches covering three orders are a disloyal workforce one better pay model away from defecting, and quality of the picked basket is the whole consumer experience. The third is agentic commerce — OpenAI already killed Instacart’s Instant Checkout once; an agent that comparison-shops across retailers by true all-in price makes Instacart’s markup structure legible and indefensible at the exact moment its catalog-data advantage was supposed to pay off.

Same playbook, new buyer

Run the picks-and-shovels play where Instacart’s marketplace conflict disqualifies it. Instacart’s enterprise stack — Storefront, Carrot Ads, Eversight, Caper — is genuinely good software sold by a vendor that also competes with its customers for the shopper relationship and the data. A pure-play enterprise version with no consumer marketplace removes the conflict grocers name explicitly, and can sell to the 38% of eGrocery that is pickup — dominated by first-party Walmart and Kroger operations Instacart never touches. The second shift is segment: Instacart Business, Restaurant Depot, and the Toast partnership hint at B2B foodservice supply, but it’s a side bet inside a consumer company; a dedicated restaurant-and-office replenishment marketplace with net-terms billing and standing orders serves a buyer with weekly recurring baskets and zero price sensitivity to delivery fees. Instacart can’t chase either wholeheartedly: its ads engine — the $1B profit center — requires consumer eyeballs on its own marketplace, so every enterprise deal that helps a retailer need the marketplace less cannibalizes the auction. The incumbent is structurally forced to protect the exact asset the new buyer doesn’t want.

Sources and further reading

• Instacart Announces First Quarter 2026 Financial Results — Instacart IR, May 6, 2026
• Instacart Reports Record Q1 With $10B GTV, $1B Revenue — Progressive Grocer, May 2026
• Instacart reports mixed Q4 results — Grocery Dive, Feb 2026
• Instacart names Chris Rogers as CEO after Fidji Simo’s exit for OpenAI — CNBC, May 28, 2025
• Instacart founder launched 20 failed companies — now he’s a billionaire — CNBC, Sept 2023
• Instacart’s down-round IPO — Axios, Sept 19, 2023
• Instacart revenue, valuation & funding — Sacra, 2026
• Instacart’s AI-Enabled Pricing Experiments May Be Inflating Your Grocery Bill — Consumer Reports / Groundwork Collaborative, Feb 2026
• July 2025 U.S. eGrocery Sales Jump to $10.0 Billion — Brick Meets Click, Aug 2025
• Could Groceries Be DoorDash’s Next Big Profit Engine? — The Motley Fool, Feb 20, 2026
• Instacart and OpenAI partner on AI shopping experiences — OpenAI, Dec 2025