Maplebear Inc. (NASDAQ: CART) — A Cash-Rich Grocery Marketplace Whose Whole Thesis Rests on an Ad Rate That Hasn’t Budged in Three Years

Independent Equity Research | Report date: 2026-07-10

Company: Maplebear Inc., doing business as Instacart. Exchange: NASDAQ. Ticker: CART. Sector: Consumer Discretionary — Internet Retail / Grocery Technology & Delivery. Price used throughout: ~$48.28 (2026-07-09 close). Market cap ~$11.2B; enterprise value ~$10.8B (net cash ~$0.63B).

⚡ The Author’s Take

This block is the author’s own independent opinion and general information — not investment advice. It is the single place a directional view or valuation zone appears; the analytical body that follows deliberately carries no recommendation and no price target.

Verdict: HOLD / show-me — a genuinely profitable, net-cash, free-cash-generative business at a fair (not cheap) price, whose upside is a real but unproven option on breaking a three-year advertising plateau. Not a short. I’d get constructive on weakness into roughly $36–42 (≈8–9x adjusted EBITDA, ≈10x FCF, a mid-single-digit and rising FCF yield) — a price at which you are paid to wait for the ad-rate inflection rather than paying up for it. Conviction: medium.

The one-liner: the arms dealer of online grocery — cheaper than DoorDash for good reasons, priced for an ad engine it keeps promising to re-light. Instacart is two businesses wearing one ticker. The first — a low-teens-growth grocery delivery marketplace, 72% of revenue — has no durable moat: the 10-K itself concedes consumer switching costs are “low,” every side multi-homes, and its ~21–22% online-grocery share is flat, defended by cutting basket prices harder than DoorDash or Uber. The second — a retail-media (advertising) + enterprise-tech engine, 28% of revenue but the large majority of the profit — is genuinely good: capital-light, ~near-100% incremental margin, built on 1.6 billion orders of proprietary grocery purchase data and deep “land-and-expand” integrations into 380+ retailer storefronts. The entire bull case is the migration from business one to business two. The problem: the advertising investment rate has been stuck at ~2.9% of GTV for three straight years (2.87% → 2.86% → 2.86%), even as management dangles a 4–5% target that would roughly double ad revenue on flat volume. Until that plateau cracks, you are underwriting a hope against Amazon (79.7% of US retail media) and Walmart.

Framing: a show-me / quality-at-a-fair-price story with an embedded AI option that cuts both ways — not deep value, not momentum, not a falling knife. The factor tape agrees: this is a low-beta (0.77), no-style-identity, ~83%-idiosyncratic name that has round-tripped a wide $33–$53 range and sits in the upper half of it after a constructive but young ~47% recovery off the February-2026 low. The quality is real (14% ROIC, ~$0.9B FCF, fortress balance sheet, single-class governance — rare founder-tech restraint). But the valuation already pays for a competent execution story, the founder-visionary (Fidji Simo) has left for OpenAI, and the buyback is only ~46% net-accretive after the SBC treadmill. What flips me bullish: the ad investment rate breaking decisively above ~3.2–3.5% of GTV for two-plus quarters — proof the retail-media engine can outgrow GTV structurally, plus evidence the ChatGPT/Claude agentic channels are net-accretive fulfillment demand rather than disintermediation. What flips me bearish: GTV decelerating toward mid-single-digits while ad growth stays stuck near 10% and fee cuts deepen — confirmation the marketplace is commoditizing and the ad ceiling is near.

📈 Stock Price Action — Five-Year Event Map

Instacart has only traded since its September-2023 IPO, so this is the full ~2.8-year public record, not a five-year map. Maplebear priced its IPO at $30.00, opened at $42.00 on 2023-09-19, and closed its first day at $33.70 before sliding below the IPO price within days. The stock bottomed near $22 (January 2024), round-tripped to an all-time high of $53.44 (2025-02-19), then chopped between the low-$30s and low-$50s for eighteen months on its own earnings prints and the AI-shopping narrative. At $48.28 (2026-07-09) it sits in the upper half of that range — roughly −6.7% below its 52-week closing high (~$51.77, Aug-2025) and −9.6% off the all-time high; 52-week range ~$32.93–$51.77. Net: it has more than doubled off the IPO-era low, but the underlying pattern is a wide, idiosyncratic trading range it is now re-testing from above. Price levels are FACT (AZI price history); attributed drivers are INTERPRETATION unless a filing or print is cited.

Cycle narrative. (1) IPO slump — priced at $30 into a cold tape, popped to $42 intraday, closed $33.70, broke IPO price within a week; gig-economy skepticism and no proven bottom line drove it to ~$22 by January 2024. (2) First-profitability re-rating — the 2024-02-13 Q4’23 report (first year of positive adjusted EBITDA, scaling advertising) flipped the narrative to “profitable ads-and-marketplace,” a ~+70% run. (3) Give-back — spring/summer 2024 prints showed order/GTV deceleration; the 2024-08-05 risk-off pushed it to ~$30. (4) Run to the ATH — reaccelerating GTV and high-margin ad growth carried it to $53.44 on 2025-02-19. (5) Earnings de-rate — the 2025-02-25 Q4’24 report paired good results with a soft Q1’25 guide; the stock fell ~30%. (6) Recovery + smooth CEO handoff — a +14% Q1’25 pop and a Q2’25 beat rebuilt it to ~$51; the Simo→OpenAI departure and Rogers promotion (announced 2025-05-28, effective 2025-08-15) were absorbed without disruption. (7) AI-disintermediation scare — OpenAI’s late-September-2025 ChatGPT “Instant Checkout” rollout sparked fears that shopping agents disintermediate Instacart’s app; ~−21% to ~$37. (8) Base-building + ad reacceleration — the 2026-02-12 Q4’25 print (GTV growth to 14%, ads over $1B) marked the ~$33 low; Instacart then turned the ChatGPT threat into a partnership (its own Instant-Checkout app inside ChatGPT, December 2025) and recovered ~+47% to $48.28.

1. Executive Summary

Maplebear Inc., which operates as Instacart, is the largest independent online-grocery marketplace in North America and, increasingly, a retail-media network and enterprise-technology vendor to grocers. In FY2025 it processed $37.2B of Gross Transaction Value (GTV) across 338.8M orders and 2,200+ retail banners, converting that into $3,742M of revenue (a ~10% “take rate”), $1,087M of adjusted EBITDA (2.9% of GTV), $447M of GAAP net income, and ~$910M of free cash flow. It is net-cash (~$0.63B), essentially debt-free, and has repurchased ~$2.76B of stock over two years.

The analytical core of this company is a two-speed split. Transaction revenue ($2,677M, 72% of the total) is the low-margin, low-moat delivery marketplace: it grows with GTV (~11%), carries thin ~7.4%-of-GTV gross economics, and sits in a structurally poor profit pool contested by Walmart (#1 online grocery, own logistics), Amazon (vertical integration), and DoorDash/Uber (superior local density). Advertising & other revenue ($1,065M, 28% of the total) is the opposite: capital-light, near-100% incremental margin, built on 1.6 billion orders of first-party grocery data and deep enterprise integrations — the engine that produces most of the profit. The whole investment case is whether Instacart can keep migrating its economics from the first business to the second.

The bull case is a capital-light flywheel with a clear runway: advertising at 2.9% of GTV today versus a 4–5% target, which would roughly double ad revenue on flat volume, plus a defensible “arms-dealer” position (Storefront Pro, Carrot Ads, Caper smart carts) as the e-commerce and ads backbone for every grocer except Walmart and Amazon. The bear case is that the ad investment rate has been flat at ~2.9% for three straight years while Amazon (79.7%) and Walmart (8.0%) absorb ~89% of incremental US retail-media dollars, leaving Instacart a ~1.6% sliver; that the marketplace is a share-flat, low-switching-cost business now cutting prices defensively; and that agentic commerce (ChatGPT, Claude) threatens the consumer relationship the data moat depends on.

On the numbers the franchise is higher-quality than the headline “gig delivery” label implies — 14% ROIC, real FCF, a fortress balance sheet, and (unusually for founder-era tech) a single class of stock with one vote per share. But the quality carries caveats: SBC is still ~9% of revenue and the much-touted buyback is only ~46% net-accretive after offsetting employee equity; the founder-visionary CEO left for OpenAI in 2025; and executive pay had no performance metrics through 2025 (fixed only after a ~55% say-on-pay vote). Valuation is fair, not cheap — fully re-rated off the 2023-24 trough (EV/sales ~0.3x → ~2.8x) but a defensible discount to DoorDash (~0.29x vs ~0.7x EV/GTV; ~10x vs ~30x+ EV/adjusted-EBITDA) earned by slower growth and grocery’s thinner take-rate. The dominant swing variable is not a factor regime but the advertising-rate inflection and the AI-agent narrative. This memo takes no position and sets no price target; the labeled opinion block above is the single exception.

2. Business Overview

What it is. Maplebear Inc. (dba Instacart), incorporated in Delaware, headquartered in San Francisco, describes itself as a “diversified technology business” operating a platform that connects four constituencies: (1) retailers (grocers and other merchants), (2) customers/end users (households ordering groceries), (3) advertisers/brands (CPG companies buying ad placement), and (4) shoppers (independent-contractor gig workers who pick and deliver orders). Operations are “largely throughout the United States and Canada”; international is only ~$46M of the ~$1,019M Q1’26 revenue (~4.5%), and no single foreign country reaches 10% of revenue.

How it makes money — two revenue lines (FACT, FY2025 10-K):

• Transaction revenue — $2,677M (71.5% of revenue). Fees earned from (a) retailers — service fees for connecting them to customers on the Instacart Marketplace and fees for orders processed on retailers’ own websites/apps powered by Instacart’s enterprise “Storefront” technology — and (b) customers — delivery and service fees plus Instacart+ subscription fees. Critically, transaction revenue is recognized net of the cost of goods remitted to retailers, shopper pay, coupons, incentives, and refunds. This is why revenue ($3.7B) is a small ~10% slice of the $37.2B of GTV that flows through the platform: Instacart is an agent, not a reseller of groceries.

• Advertising & other revenue — $1,065M (28.5% of revenue). The sale of advertising services to brands — principally Sponsored Product ads (the main growth driver), plus display ads, coupons, and brand pages — recognized as clicks/impressions are delivered, plus an “other” bucket of enterprise SaaS fees. Brand advertising contracts are typically under one year with no long-term commitments — a structurally lower-visibility revenue stream than subscription software, though sticky in practice for large CPGs.

The GTV → revenue → take-rate structure. GTV is the total value of products sold through Instacart, including taxes, deposits/fees, customer tips (which pass through to shoppers), Instacart+ fees, and other fees. The relationships that matter:

(FACT, FY2023/2024/2025 10-Ks. 2022 transaction revenue split approximate.) Two structural signals fall out of this table and recur throughout the memo: (1) the take-rate expansion is over — it climbed from 7.4% to ~10% through 2023 as Instacart layered fees and advertising onto GTV, then plateaued at ~10.0–10.1%, so revenue now grows roughly in line with GTV; and (2) the advertising investment rate (ads/GTV) has been stuck at ~2.9% for three years — the single most important number in the entire thesis.

The enterprise “arms-dealer” stack. Beyond the consumer app, Instacart sells a modular B2B technology suite that retailers adopt à la carte — the strategic through-line of essentially all of its M&A: Storefront / Storefront Pro (white-label e-commerce powering 380+ grocery sites including Costco, Publix, Sprouts, ALDI; management claims upgraders see “+10pp online-sales lift, +5pp 90-day new-user retention”); Fulfillment API (Instacart shoppers or the retailer’s own employees pick/pack/deliver); Connected Stores in-store tech — Caper Carts (AI smart carts, live in 100+ cities), FoodStorm (catering order management), Carrot Tags (electronic shelf labels), Store View (computer-vision shelf-availability); Carrot Ads (retail-media technology extended to retailers’ own storefronts, 310+ partners); Insights (a first-party data portal subscribed to by Kraft Heinz, Advantage Solutions); and a new AI Solutions / Cart Assistant (Kroger and Sprouts pilots). This layer is where the genuine switching costs and the durable part of the business live.

Instacart+ membership ($99/year) offers unlimited $0 delivery over a minimum basket (cut to $10 for grocery in 2025). Management states members are “the majority of GTV and orders,” order roughly twice as often as non-members, and retain materially better; deferred revenue on the balance sheet ($275M) is primarily Instacart+. Membership is Instacart’s main captivity/habit mechanism — real, but structurally weaker than Amazon Prime because it is single-purpose and undercut by low switching costs.

Verdict. A capital-light, agency-model marketplace with a genuinely valuable second act (advertising + enterprise tech) bolted onto a thin, low-margin transaction core. The business is more profitable and higher-quality than the “gig delivery” caricature, but its economics are only as good as its ability to keep shifting weight onto the advertising and enterprise layers — which the flat take-rate and flat ad-rate say has stalled at the margin.

3. Industry Dynamics

A large, under-penetrated, fast-growing category. US online grocery reached roughly 19% of total grocery spend at YE2025 (up ~430bps year-over-year), with monthly eGrocery sales around $12.3–12.5B growing ~29–31% year-over-year in peak months; delivery is ~45% of eGrocery spend and growing ~30–36%; household penetration reached ~82.9M US households (+12.9%) (FACT, Brick Meets Click / Mercatus 2025). Grocery is the largest retail category and among the least e-commerce-penetrated (~15–19% vs ~30%+ for general merchandise), so the structural online migration is a genuine multi-year runway. This category tailwind is the strongest single leg of the bull case at the industry level.

But the profit pools split sharply — this is a barbell industry.

• Delivery logistics is a structurally bad profit pool. Unit economics are thin to negative; gig-labor is the dominant variable cost and a regulatory tail risk; consumer switching costs are low; and the two largest players — Walmart and Amazon — own their own supply and last-mile and are willing to subsidize delivery as a loyalty/basket tool rather than run it for profit. Moving boxes in grocery is a bad business.

• Retail media (advertising) is a structurally excellent profit pool — but hyper-concentrated. US retail-media ad spend was $60.3B in 2025, growing to ~$71.1B in 2026 (+17.8%) (FACT, eMarketer), at near-100% incremental margin, as CPG trade and shopper-marketing dollars migrate to measurable first-party-data networks amid third-party-cookie deprecation. The catch: Amazon Ads is ~79.7% of US retail media and Walmart Connect ~8.0%, and the two are forecast to capture ~89% of incremental spend in 2026. Instacart and DoorDash each run ~$1B of annual US ad revenue — ~1.6% share apiece; Uber Advertising is ~$1.5B (+60%). Instacart plays in the good pool, but as a low-single-digit-share niche behind two giants.

Market share — strong among independents, contested overall. On a total-online-grocery basis, 2025 share is roughly Walmart ~25.7% (#1), Amazon ~22.0% (#2), Instacart ~21.6% (#3) — and Instacart’s share is roughly flat, not gaining. On a third-party-delivery-only view Instacart is a clear leader (~29%), and among grocery-delivery platforms (Modern Retail) DoorDash ~⅓, Uber ~20%, Instacart ~20%. So Instacart is a strong #1 among independent grocery-delivery marketplaces but a contested #3 in the broader market where Walmart and Amazon own their own logistics.

Competitor momentum. DoorDash grocery is at a >$7B GTV run-rate with “new verticals” growing faster than restaurants (though unit-economic-negative through 2025); Uber’s non-restaurant delivery is a ~$12.5B gross-bookings run-rate (+25%), grocery ~18% of Uber delivery, adding 1,000+ retailers in 2025; Walmart+ is ~$98/year; Amazon’s perishable same-day sales reportedly grew ~40x since early 2025 (though Amazon’s total US grocery share remains <4%, its weakest retail category).

Capital-cycle read (Marathon lens). Online grocery/quick-commerce attracted enormous capital in 2020–21 (Instacart, DoorDash, Uber, Gopuff, Getir, Jokr, dark stores). The 2022–24 shakeout rationalized the fringe (Getir collapsed, Gopuff retrenched, the dark-store model largely failed) and the survivors turned profitable — the favorable post-bust phase Marathon prizes. But unlike a clean consolidation, grocery specifically is re-intensifying: DoorDash and Uber are plowing rationalized-core cash into grocery “new verticals,” and Walmart/Amazon are structurally advantaged and subsidizing. That is Marathon’s warning pattern — cash cows funding capital-hungry adjacencies aimed at the same pool. Instacart’s strategic answer is to exit the head-on capital war and become the neutral arms-dealer (Enterprise Platform + Carrot Ads) arming every grocer except Walmart and Amazon — a picks-and-shovels position partly insulated from the delivery-subsidy war.

Verdict: structurally mixed / barbell. A good industry to own data and relationships in; a bad industry to move boxes in. The category migration is real and durable, but the delivery-logistics pool where Instacart’s marketplace primarily sits is structurally poor, while the retail-media pool it is pivoting toward is excellent yet ~88% owned by two giants. Instacart’s viability rests on migrating its economics from the bad pool to the good one — which is exactly what the ads/enterprise pivot attempts, and exactly what the flat ad-rate says it has not yet decisively achieved.

4. Competitive Position

Name the moat precisely — and concede where there isn’t one. Instacart’s consumer marketplace has no durable moat, and the 10-K says so in its own risk factors: “the cost to switch between providers of online grocery shopping is low for consumers,” who “have a propensity to shift to the lowest-cost or highest-quality provider and may use more than one platform.” Retailers multi-home (partner with Instacart and build their own storefronts and use DoorDash/Uber); shoppers are gig workers active on multiple apps. It is a four-sided platform with low captivity on every side — the textbook absence of network-effect lock-in. Where a genuine moat exists, it is two narrower mechanisms in the B2B layer:

1. Scale-based proprietary data → retail media (an intangible/scale advantage). 1.6 billion lifetime orders constitute the largest pure-play, transaction-level grocery purchase dataset outside Amazon and Walmart. It powers ad targeting and closed-loop measurement (Sponsored Products, off-platform data monetization on Meta/Google/TikTok/The Trade Desk/Pinterest, the Insights portal). This is a real advantage — but subscale: ~$1B of ad revenue is ~1.6% of US retail media against Amazon’s 79.7%. It yields a narrow, high-margin niche, not category dominance.

2. Enterprise switching costs (the under-appreciated durable moat). Storefront Pro + Carrot Ads + Caper + FoodStorm + the fulfillment API create deep, multi-year technical integrations into a retailer’s e-commerce, catalog, search, checkout, in-store hardware, and ad stack. 380+ storefront clients, a “land-and-expand” motion (adopt Storefront → add Carrot Ads → add Caper → add AI Solutions), with ALDI, Costco, Publix, and Sprouts deeply embedded. The CEO’s framing — “the breadth and depth of our platform is difficult to replicate” — is defensible here: this is a genuine supply-side + customer-captivity moat with real switching costs, arguably more durable than the consumer marketplace, and it feeds the data/ads moat (each Storefront client “almost always enables Carrot Ads”).

A third candidate — local shopper density/liquidity — is weak. DoorDash (~65% US restaurant share) and Uber have greater raw courier density; Instacart’s edge is grocery-specific pick expertise (in-store shoppers who understand produce and substitutions), not delivery density.

Share-stability test (Greenwald). A genuine wide moat gains share or defends high stable share at high ROIC. Instacart’s online-grocery share is roughly flat (~21–22% total; ~29% of 3P delivery) — and it is being defended with defensive price/fee cuts (Wells Fargo channel work: Instacart cutting basket prices ~6% vs DoorDash ~2%, Uber ~1%). Flat share in a fast-growing category, held via price cuts, is the signature of a narrow moat under pressure, not a widening one.

Head-to-head.

• vs. DoorDash (grocery): superior local density and a larger membership base (35M+), grocery at a >$7B GTV run-rate — but its grocery new verticals were unit-economic-negative through 2025, and DoorDash’s own 10-K concedes it competes with “Instacart’s retailer relationships.” Instacart’s edge: deeper retailer integrations, larger baskets, grocery-native pick quality.
• vs. Uber Eats: partner and competitor — Uber powers Instacart’s restaurant tab, yet competes in grocery delivery and advertising (~$1.5B run-rate).
• vs. Walmart (Walmart+ / in-house): the #1 online grocer owns supply + logistics + a ~$5B+ ad network and can subsidize delivery — a structural cost/scale advantage Instacart cannot match on price.
• vs. Amazon (Fresh/Whole Foods + Amazon Ads): vertically integrated, 79.7% of retail media, deepest balance sheet — but <4% grocery share (grocery is Amazon’s weakest category).
• vs. retailers’ own apps (the existential swing): Instacart’s enterprise moat depends on retailers choosing to outsource their e-commerce and ads. The value prop (scale, +10pp sales lift) holds for mid-size grocers, but large retailers can build in-house — Kroger ended Instacart exclusivity and moved delivery volume toward DoorDash.

Agentic-commerce disintermediation (the key forward moat threat). If consumers shop via AI agents (ChatGPT, Claude, Gemini) that choose the fulfillment provider, the consumer relationship and first-party data — the foundation of both moats — could be disintermediated. Management’s answer: be “wherever customers are while maintaining control of the experience and our data,” build “the gold standard of agentic grocery AI” (Cart Assistant, reportedly reaching ~25% of US customers), and treat third-party AI as “an incremental demand channel” (Instacart became the first app with embedded Instant Checkout inside ChatGPT in December 2025, and per the CEO integrated “most recently with Claude”). This is a genuine, unquantifiable tail risk; the mitigation is credible but unproven, and Instacart is a taker, not a setter, of the agentic-commerce rules.

Verdict: narrow, bifurcated moat — real but contested. The consumer marketplace is a low-switching-cost, multi-homed, share-flat delivery business with no durable moat. The genuine, defensible moat is the B2B layer: proprietary grocery data (feeding a subscale-but-high-margin retail-media network) plus deep enterprise switching costs (Storefront Pro / Carrot Ads / Caper “land-and-expand”). That moat is durable against smaller players and against mid-tier grocers’ build-vs-buy, but it is subscale versus Amazon/Walmart in the profit pool that matters and exposed to agentic disintermediation of the consumer relationship it rests on. A narrow moat migrating from a moatless delivery core to a defensible arms-dealer core — durable enough to persist, not wide enough to dominate.

5. Growth History and Forward Opportunities

GTV: durable but decelerating, and re-accelerating off a 2023 trough. GTV compounded at ~10.6% from 2021 ($24.9B) to 2025 ($37.2B), with a clear cycle: a COVID-normalization slump to +5% in 2023, then re-acceleration to +10% (2024), +11% (2025), and +13% in Q1’26 — nine consecutive quarters of double-digit YoY GTV growth per the CEO. The re-acceleration is a genuine positive; the deceleration risk is that it was partly engineered (below).

Orders vs. AOV — a quality flag. Orders grew +3% (2023) → +9% (2024) → +15% (2025) → +10% (Q1’26), while average order value drifted down ($112.6 → $113.8 → $109.9). The 2025 order surge was juiced by the $10-minimum-basket benefit for Instacart+ members, which drove many small incremental orders and depressed AOV. As Instacart laps that (Q1’26), order growth stepped down to +10% and GTV again outpaces orders. Critically, the CFO now says forward order growth will be driven primarily by user growth, with frequency “continuing to play a role” but no longer the primary lever — i.e., the cheap frequency lever ($10 basket) is largely spent, and future growth must come from harder-won new-user acquisition and basket deepening.

Advertising — the profit engine, but its key ratio is flat. Ad revenue grew $572M (2021) → $740M (+29%) → $871M (+18%) → $958M (+10%) → $1,065M (+11%), with Q1’26 re-accelerating to +16% (“fastest growth since Q3’23”) on mid-market and emerging-brand strength. But the growth rate decelerated hard from ~29% to ~10–11%, and — the crux — the advertising investment rate (ads/GTV) has been stuck at ~2.9% for three years (2.87% → 2.86% → 2.86%; 2.8% in Q1’26). Management’s stated long-term target is 4–5% of GTV, which would nearly double advertising as a share of volume. The entire high-margin bull case rests on breaking this plateau via more ad surfaces (Carrot Ads’ 310+ partners, Caper in-store aisle-aware ads), more advertisers (9,000+ brands, self-serve Ads Manager, mid-market/emerging), off-platform data monetization, AI ad formats, and new formats like the “Immersive Feed” vertical-video ads launched June 2026. The bear read: three years of a flat ~2.9% rate — while Amazon (79.7%) and Walmart absorb ~89% of incremental retail-media dollars — is strong evidence the 4–5% target is aspirational and that the plateau may be near the practical ceiling given finite grocery-basket ad load and giant competition. This ratio is the single biggest swing factor in the thesis.

Growth is overwhelmingly organic. M&A is small tuck-in B2B tech (Caper ~$350M in 2021; FoodStorm/Eversight/Rosie undisclosed 2021-22; Wynshop ~$106M 2025; Instaleap 2026) — capability bolt-ons for the enterprise stack, not GTV-buying. FY2025’s $106M of acquisition spend is immaterial against $37B of GTV. There is no roll-up masking here — a genuine quality positive.

Forward drivers (transcript). (1) Advertising penetration 2.9% → 4–5% of GTV (the big one); (2) enterprise expansion — 380+ storefronts, land-and-expand; (3) in-store tech — Caper Carts (grocery is still ~85%+ in-store, so the TAM is the physical basket); (4) international — enterprise-led (Storefront Pro for Costco Spain/France “ahead of expectations,” Instaleap for LatAm/Europe fulfillment), early and disciplined; (5) restaurants (via Uber, low-margin); (6) AI — Cart Assistant, agentic commerce, better search/personalization. Each lever is either subscale (international, in-store ads), low-margin (restaurants), or dependent on the unproven ad-rate break; the collective reads as “durable low-teens GTV plus slightly faster revenue,” not hyper-growth. Q2’26 guidance codifies this: GTV $10.1–10.25B (+11–13%), advertising +11–14%, adjusted EBITDA $290–300M (+11–15%).

Verdict: mixed-quality, durable-but-decelerating growth. The marketplace (72% of revenue) is a mature, low-teens, low-quality, competitive volume business whose best frequency lever is spent. The advertising engine (28% of revenue, most of the profit) is genuinely high-quality — organic, high-margin, capital-light — but its key ratio has been flat for three years, making the 4–5%-of-GTV target, the fulcrum of the bull case, unproven. Growth is real, organic, and durable at low-teens GTV; it is not clearly high-quality in aggregate because the profitable half depends on breaking a three-year plateau against dominant competition.

6. Financial Quality

Revenue and its composition. Revenue grew $1,834M (2021) → $3,742M (2025), ~+11% in each of the last two years, split into the low-margin transaction line (72%, tracking GTV) and the high-margin advertising line (28%, near-100% incremental margin and reaccelerating to +16% in Q1’26). Advertising is ~28% of revenue but a materially larger share of profit — the correct way to read the P&L.

Margins and operating leverage. Gross margin is ~74% of revenue (gross profit is ~7.4% of GTV, stable). GAAP operating margin was 13.3% in FY2025 — a swing from −70% in 2023, when the IPO triggered a one-time ~$2.76B stock-comp vesting charge. Adjusted EBITDA grew $641M → $885M → $1,087M (2.1% → 2.6% → 2.9% of GTV) — genuine operating leverage as the ads/enterprise mix scaled on a roughly fixed cost base. Q1’26 adjusted EBITDA was $300M (+23%).

Quality of earnings — the SBC caveat. The gap between adjusted EBITDA ($1,087M) and GAAP EBITDA ($597M) is mostly stock-based compensation ($352M in FY2025, ~9.4% of revenue). SBC has fallen sharply from the $2,756M IPO-vesting spike (2023) but remains high for a business management frames as “mature,” and it absorbs ~39% of free cash flow via the buyback needed to offset dilution. GAAP diluted EPS was $1.57 in FY2025; note that the FY2022 “net income” of $428M was a deferred-tax valuation-allowance release, not operations. Net income is only ~1.2% of GTV — a reminder of how thin the underlying take is.

Cash generation. Operating cash flow was $971M in FY2025 against just $61M of capex (1.6% of revenue — genuinely asset-light), for ~$910M of free cash flow (24% of revenue; ~84% of adjusted EBITDA). FCF exceeds net income (non-cash SBC and working-capital timing). This is real cash, not an accounting artifact.

Balance sheet — fortress. Net cash ~$0.63B (cash + short-term investments ~$0.69B against ~$34M of capital-lease debt), a new undrawn $500M revolver, current ratio 2.36x, no pension, minimal operating leases. The ~$1.1B receivables balance is retailer float (Instacart collects from consumers and remits to retailers). Book equity ($2.4B) sits against a −$4.7B accumulated deficit (IPO SBC) and $7.1B of paid-in capital — which is why P/B is a meaningless lens here.

Returns. ROIC ~14%, ROA ~11.5%, ROE ~18–19% — but ROE flatters on a buyback-depleted equity base; ROIC and FCF are the right return lenses, and both are respectable for an asset-light platform.

Verdict: economics do improve with scale — with a real blemish. The model is asset-light, high-gross-margin, advertising-led operating leverage with strong FCF conversion and a fortress balance sheet. The blemishes: SBC at ~9% of revenue dilutes the GAAP picture and funds a buyback treadmill, and the marketplace itself is a thin (~2–3% of GTV) profit business whose economics rest on the advertising overlay. Higher-quality than the “gig delivery” label; not as clean as the adjusted numbers imply.

7. Capital Allocation

The cash-return story — and the math that deflates half of it. Instacart has repurchased ~$2.76B of stock over two years ($1,407M FY2024 + ~$1,349M FY2025) — headline-aggressive for an ~$11B company, funded entirely from FCF and net cash, with no dividend (and none intended). But the net effect is far smaller than the gross: shares outstanding fell from 279M (YE2023) → 261M (YE2024) → 243M (YE2025) → 237M (Q1’26) — a net reduction of ~36M shares against ~78.8M shares gross-repurchased. In other words, ~42.8M shares were re-issued via SBC/RSU settlement, so only ~46% of the gross buyback became real share shrink; ~54% simply offset employee-equity dilution. Management deserves less credit than the $2.8B headline implies — the “cannibal” is roughly half a treadmill. Additionally, ~$507M (~18% of the two-year buyback) went to privately negotiated repurchases from pre-IPO/venture/related-party holders (Feb-2024 ~$390M at ~$27, Aug-2024 ~$117M at ~$32) — useful for absorbing insider overhang, but a managed insider-exit vehicle at below-current prices, not pure open-market accretion. At YE2025 ~$671M remained authorized ($~323M after Q1’26’s $349M).

M&A — disciplined and strategic. Every acquisition has been a small tuck-in B2B tech capability, not a GTV grab: Caper AI (smart carts, ~$350M, 2021 — the only disclosed price), FoodStorm/Eversight/Rosie (undisclosed, 2021-22), Wynshop (grocery e-commerce storefront/fulfillment SaaS, ~$106M, 2025), and Instaleap (LatAm/Europe fulfillment platform, 2026 — first international push). Goodwill ($393M) and intangibles ($453M) are modest; there is no scale-goodwill-impairment risk and no serial-acquirer red flag. The through-line is coherent: build out the enterprise/retail-tech stack (OMS, carts, ad/pricing, storefronts, international fulfillment) — the durable side of the business.

Governance — a genuine positive. Unusually for founder-era tech, Instacart has a single class of common stock with one vote per share — no super-voting, no dual-class. That materially raises governance quality versus DoorDash (founder-absolutist control) and most 2020-23 tech IPOs, and it means the buyback and board are accountable to economic owners.

Incentive alignment — historically weak, improving under pressure. Executive pay has been “weighted heavily toward long-term equity” but primarily time-vesting RSUs; the performance-based Executive Bonus Plan was not used for NEOs in FY2024 or FY2025 — meaning essentially no GTV / adjusted-EBITDA / TSR metric gated executive pay through 2025. Say-on-pay drew only ~55% support at the 2025 AGM; in response, management engaged its 14 largest holders (>47% of shares) and is introducing relative-TSR PSUs for CEO Rogers’ 2026 refresh plus “boxcar” (back-loaded) vesting. FY2025 NEO comp: Chris Rogers (CEO) $29.85M, Emily Reuter (CFO) $13.67M, Morgan Fong (CLO) $12.49M. Former CEO Fidji Simo forfeited all unvested RSUs and her 2022 PSU on her 2025-08-15 resignation. Ownership guidelines (CEO 5x salary), a clawback policy, and a hedging/pledging prohibition are in place. Net: alignment is improving, but only after an investor rebuke.

Insider signal — net distribution, no conviction. No open-market discretionary purchases (code P) appear in the Form 4 corpus; officer filings are routine tax-withholding on vesting (code F). Venture/director selling is the pattern — Sequoia partner Ravi Gupta (director) sold 181,000 shares at $41.51 (June-2026); Sequoia filed a 13D/A in June-2025. This is the neutral-to-mildly-negative signal typical of a post-IPO venture unwind, not a red flag — but there is no insider vote of confidence via open-market buying. Key holders (2026 proxy): Sequoia ~12%, D1 Capital ~11% (both with board seats), co-founder Apoorva Mehta ~9%, Vanguard ~8%, BlackRock ~6%; all execs/directors ~24%; large public float. Saudi Arabia’s PIF, a pre-IPO holder, has dropped below the 5% table.

Verdict: above-average for a 2023-vintage IPO, with real caveats. Positives: genuine net cash + ~$0.9B FCF, disciplined tuck-in M&A, single-class governance, real (if partial) share shrink, absorbed venture overhang. Negatives: the headline ~$2.8B buyback is only ~46% net-accretive; SBC is still ~9% of revenue; ~$507M funneled to insider exits below current prices; historically weak pay-for-performance fixed only after a ~55% say-on-pay vote. The bull’s “disciplined net-cash cannibal” is half right; the bear’s “buyback masks dilution” is half right; the truth is in the middle and quantifiable.

8. Changes and Headwinds — Last Two Years

CEO transition — the biggest single change (narrative-negative, execution-neutral). On 2025-05-08 founder-era CEO Fidji Simo announced her departure to become OpenAI’s “CEO of Applications” (stock ~−4% intraday). The board named Chris Rogers — previously Chief Business Officer and the architect of the ads/enterprise engine — CEO effective 2025-08-15. Simo resigned as board Chair in November 2025; on 2026-07-09 she stepped back from her OpenAI No.2 role to a part-time advisory capacity (medical leave). The read: a loss of the founder-visionary halo/leadership premium, offset by continuity in the profitable ads/enterprise pivot (Rogers built it) and retention of the OpenAI/ChatGPT channel. Roughly execution-neutral, narrative-negative.

Competitive escalation + defensive price/fee cuts — the core bear catalyst. DoorDash and Uber pushed hard into grocery; Kroger ended Instacart exclusivity and moved delivery volume toward DoorDash; ad growth decelerated to ~10–11% with the investment rate plateauing at ~2.9% of GTV; and by mid-2026 Instacart was cutting product prices and fees more than rivals (basket ~−6% vs DoorDash ~−2%, Uber ~−1% per Wells Fargo) — a direct margin and growth pressure that is the clearest bear catalyst.

Strategic/product changes. The Uber Eats partnership (2024) added a restaurant tab powered by Uber (Instacart cedes restaurant fulfillment economics but plugs a demand gap); the $10-minimum-basket affordability push (2025) drove order growth but depressed AOV; the AI/agentic pivot — Cart Assistant and AI Solutions (Nov-2025, Kroger/Sprouts pilots), and Instacart becoming the first app inside OpenAI ChatGPT with embedded Instant Checkout (Dec-2025) and integrating with Claude — turned the AI-disintermediation threat into a demand channel. Emily Reuter was promoted to CFO (May-2024).

Regulatory / one-time items. A San Diego City Attorney worker-misclassification settlement of $46.5M reached final judgment in 2024 (no admission); California Prop 22 was upheld by the state Supreme Court in July-2024 (favorable — preserves the independent-contractor model). Instacart paid ~$60M in regulatory settlements in Q1’26 (a cash-flow item), and a Q1’26 benefit from the Canada digital-services-tax repeal is non-recurring. An FTC pricing-tool probe was reported in December-2025. Gig-worker reclassification remains the structural regulatory tail.

Verdict: net mildly weakens the thesis. The two biggest changes — the CEO handoff and the competitive escalation / defensive price-fee cuts — are, on balance, modestly negative (loss of founder narrative; direct margin pressure), partly offset by the AI/agentic optionality and the resilient Q1’26 beat (record GTV $10.29B, adjusted EBITDA +23%). None is thesis-breaking; together they argue for “durable but pressured,” not “inflecting.”

9. Risk Analysis (Risk Matrix)

Catastrophic-loss risk is low. Net cash, ~$0.9B FCF, asset-light, no leverage, defensive category — the balance sheet all but rules out a solvency event. The realistic downside is a valuation de-rate on a growth/ad-rate disappointment, not a wipeout.

10. Valuation Discussion (Embedded Expectations)

At $48.28, market cap is ~$11.2B and enterprise value ~$10.8B basic (~$12.2B on ~265M diluted shares — call it ~$11.3B), net of ~$0.63B net cash. The multiples, with the essential distinction between GAAP and company-adjusted EBITDA drawn explicitly:

Read the P/B percentile carefully — and then ignore it. AZI flags P/B at the 97.5th percentile (“richest ever”), but book equity is shrinking mechanically (buybacks against a −$4.7B accumulated deficit), so P/B is meaningless here. The right lenses are EV/GTV, EV/adjusted-EBITDA, and P/FCF, all of which say the stock is fully re-rated off its 2023-24 trough but not egregiously expensive.

Peer frame. DoorDash trades at ~0.7x EV/GTV and ~30x+ EV/adjusted-EBITDA on ~20%+ GTV growth; Uber at ~0.2x EV/gross-bookings and ~18–22x EV/adjusted-EBITDA. Instacart’s ~0.29x EV/GTV and ~10x EV/adjusted-EBITDA is a clear discount to DoorDash — but a defensible one, earned by slower GTV growth (11–14% vs 20%+), grocery’s structurally thinner take-rate, and a weaker moat. It is not obviously mispriced in either direction.

Embedded expectations — what ~$11.3B EV / ~16x GAAP-EBITDA / ~10x adjusted-EBITDA / ~0.29x GTV requires:

• Continued double-digit GTV growth (~11–14%): the market is not pricing a stall to mid-single-digits.
• Advertising scaling from 2.86% of GTV toward ~3.3–4% — i.e., ad revenue outgrowing GTV, which is what drives EBITDA-margin expansion.
• Adjusted EBITDA/GTV rising from 2.9% toward ~3.3%+ — operating leverage holds (it went 2.1% → 2.6% → 2.9% over 2023-25).
• The AI-agent channel net-neutral-to-accretive rather than disintermediating the app. This is the single biggest swing in the multiple — the ~16x GAAP-EBITDA (vs DoorDash’s 30x+) is the AI/competition risk discount.

Three-year scenario sketch (illustrative EV zones, no price target):

• Bear: GTV fades to ~5–7% (AI agents + Walmart/Amazon/DoorDash pressure), advertising stalls at ~2.9% of GTV, adjusted-EBITDA/GTV flat → adjusted EBITDA ~$1.1–1.2B by FY28; multiple compresses to ~7–9x → EV contracts vs today. Falsifier: two-plus quarters of sub-8% GTV and flat ad growth.
• Base: GTV ~10–12%, advertising rises to ~3.3–3.5% of GTV, adjusted-EBITDA/GTV ~3.2–3.4% → adjusted EBITDA ~$1.5–1.7B by FY28; ~10–12x → EV ~$16–20B, with buybacks shrinking the share count ~3–4%/yr → mid-teens per-share compounding at a flat multiple. Falsifier: ad growth re-couples down to GTV growth.
• Bull: GTV ~13–15% (agentic commerce accretive; Instacart = fulfillment + retail-media layer), advertising ~4%+ of GTV, adjusted-EBITDA/GTV ~3.7%+ → adjusted EBITDA ~$2.0–2.3B by FY28; ~13–15x → EV ~$28–34B. Falsifier: GTV decelerates despite AI partnerships (proves disintermediation).

Verdict: roughly fair. Richer than its own 2023-24 trough, a defensible discount to DoorDash, neither cheap nor egregiously expensive — with the advertising-rate inflection and the AI-agent narrative the dominant swing factors. No price target; no recommendation (the labeled opinion block above is the single exception).

11. Variant Perception

Consensus. Sell-side and the tape treat CART as a re-rated, fairly-valued execution story — a profitable grocery-tech platform that has already round-tripped the “cheap turnaround” trade (EV/sales 0.3x in 2023 → ~2.8x now), with a credible ads/enterprise pivot and an AI-narrative overhang. It is neither a crowded momentum darling nor an abandoned value name; the factor tape confirms a low-beta (0.77), no-style-identity, ~83%-idiosyncratic “show-me” stock whose swing variable is the AI narrative, not a factor regime.

Strongest bull case. A capital-light, high-margin retail-media + enterprise flywheel with a clear runway: advertising at 2.9% of GTV against a 4–5% target would roughly double ad revenue on flat volume, dropping almost entirely to EBITDA. Instacart owns a defensible arms-dealer position — the e-commerce and ads backbone for every grocer except Walmart/Amazon (1.6B orders of data, 380+ storefronts, deepening lock-in, off-platform and in-store ad expansion) — plus real FCF, net cash, single-class governance, and a legitimate agentic-commerce option (Instacart as the fulfillment + retail-media layer for AI shopping). At ~10x adjusted EBITDA and ~12x FCF, you are not paying much for that optionality.

Strongest bear case. The advertising investment rate has been stuck at ~2.9% of GTV for three years — the 4–5% target is unproven, and Instacart is a ~1.6% sliver of a retail-media market ~80% owned by Amazon. Meanwhile the core marketplace is a low-switching-cost, multi-homed, share-flat delivery business attacked simultaneously by Walmart (#1, own logistics + scale), Amazon (vertical integration + balance sheet), and DoorDash/Uber (superior density, each growing ads faster) — forcing defensive price/fee cuts that pressure margins — while agentic commerce threatens the consumer relationship the entire data moat depends on. Strip out the ad-rate hope and you own a mid-teens-EV/EBITDA, low-teens-growth, thin-margin delivery business with a subscale ad business, at a full multiple.

The 3–5 assumptions that matter most:

1. Can the ad investment rate break ~2.9% toward 4–5% of GTV? (Bull’s fulcrum; three years of no progress.)
2. Is agentic commerce net-accretive or disintermediating to Instacart’s consumer relationship and data?
3. Does GTV hold low-teens once the $10-basket frequency lever is fully lapped and new-user growth must carry it?
4. Do retailers keep outsourcing to Instacart’s enterprise stack, or build in-house (Kroger’s path)?
5. Does Walmart/Amazon/DoorDash competition force further margin-eroding price cuts?

What would falsify each side. Bull falsified: ad revenue re-couples down to GTV growth (rate stays ~2.9%) for a year, and/or GTV decelerates toward mid-single-digits — the ads engine can’t outgrow the volume base and the moat migration failed. Bear falsified: the ad investment rate breaks decisively above ~3.2–3.5% of GTV for two-plus quarters and GTV holds low-teens with agentic channels adding incremental demand — proving the retail-media engine can structurally outgrow GTV.

12. Fact vs. Interpretation Table

13. Open Questions

1. Will the advertising investment rate break its three-year ~2.9% plateau toward 4–5% of GTV? No evidence yet that it can — the single biggest thesis swing.
2. What is the true profit split between the moatless marketplace and the moaty ads/enterprise layer? Management refuses to break out enterprise economics; the split is inferred, not disclosed.
3. How large and how fast is agentic-commerce disintermediation? Unquantifiable; watch ChatGPT/Claude channel mix and whether Instacart retains first-party-data control.
4. How durable are the large retailer relationships (top-3 ~43% of GTV) given Kroger’s move toward DoorDash and large grocers’ build-vs-buy incentive?
5. Where does GTV growth settle once the $10-basket frequency lever is fully lapped and growth must come from new users?
6. What is the residual Saudi PIF stake and the remaining venture (Sequoia/D1) overhang to be worked off via secondaries/buybacks?

14. What Must Be True

Bull case — what must be true: The advertising engine must structurally outgrow GTV — the ad investment rate breaks above ~3.2–3.5% of GTV and marches toward 4–5%, driven by off-platform data monetization, in-store/Caper ad surfaces, and AI ad formats — and GTV holds low-teens as the enterprise arms-dealer position deepens and agentic channels add incremental demand. If so, adjusted EBITDA compounds toward ~$1.7–2.3B by FY28 on high-incremental-margin ad dollars, and a net-cash, share-shrinking, single-class-governed platform re-rates or compounds per-share value at a healthy clip.

Falsification test: the ad investment rate stays at ~2.9% of GTV (ad revenue growth re-couples down to GTV growth) for a full year — proof the retail-media engine has hit a practical ceiling against Amazon/Walmart and cannot carry the story.

Bear case — what must be true: The marketplace commoditizes — Walmart/Amazon/DoorDash/Uber competition forces continued defensive price/fee cuts, GTV decelerates toward mid-single-digits as the frequency lever laps, the ad rate stays capped at ~2.9%, and agentic commerce begins to disintermediate the consumer relationship and first-party data. If so, adjusted EBITDA stagnates near ~$1.1–1.2B, the ~16x GAAP-EBITDA multiple compresses toward the Uber band, and the stock de-rates.

Falsification test: the ad investment rate breaks decisively above ~3.2–3.5% of GTV for two-plus consecutive quarters while GTV holds low-teens — proof the moat migration is working and the marketplace is not commoditizing fast enough to matter.

The single most important variable for both sides is the advertising-investment-rate trajectory. Everything else — moat durability, competitive pressure, valuation — resolves around whether that ratio breaks its three-year plateau.

15. Source Appendix

See the Source Appendix below for the full list of primary and secondary sources, with URLs and access dates. Primary sources relied on include: Maplebear Inc. FY2023/FY2024/FY2025 Forms 10-K; the Q1’26 Form 10-Q; the 2026 DEF 14A proxy; 8-K material-event filings (CEO transition, quarterly results); Form 4 insider filings; the Q1’26 earnings-call transcript; ROIC.ai fundamentals/valuation data; the AZI price history and valuation-percentile feed; the FactorsToday factor model; and third-party industry data (Brick Meets Click / Mercatus on eGrocery penetration; eMarketer on retail media).

This article is independent research and general information, not investment advice. The analytical body carries no recommendation and no price target; the author’s directional view appears only in the labeled opinion block at the top.

APPENDIX A — Standard Diligence Questionnaire

Maplebear Inc. (NASDAQ: CART) — dba Instacart | Report date: 2026-07-10

Supplemental to the research memo. Fact / Interpretation / Assumption labels applied where it matters.

General

What thoughtful questions have other investors asked about this company? The recurring institutional questions cluster on: (1) Can advertising break its ~2.9%-of-GTV plateau toward the 4–5% target? — the crux of the model; (2) Is Instacart structurally disadvantaged versus Walmart/Amazon (who own supply + logistics and subsidize delivery) and losing relevance to DoorDash/Uber’s density; (3) Does agentic commerce (ChatGPT/Claude) disintermediate the consumer relationship and the first-party data moat?; (4) What is the real net-share-count trajectory given SBC at ~9% of revenue offsetting the buyback; and (5) How durable are the large retailer relationships (top-3 ~43% of GTV) after Kroger ended exclusivity.

Cyclicality & Earnings Nature

Cyclical high or low? Neither extreme. Grocery is defensive/staple demand, so GTV is relatively acyclical; margins are at a structural early-expansion phase (adjusted-EBITDA/GTV 2.1% → 2.9% over 2023-25) rather than a cyclical peak. Interpretation: earnings are closer to an early-normalization than a cyclical top or trough — the swing is competitive/strategic (ad rate, price cuts), not macro.

Driven by external environment or internal actions? Predominantly internal — advertising monetization, enterprise expansion, cost discipline, and buybacks. External factors (online-grocery penetration ~19% and rising; retail-media secular growth) are tailwinds; competitive intensity (Walmart/Amazon/DoorDash/Uber) is the main external headwind.

How stable are revenues? Transaction revenue (72%) is stable and recurring in aggregate (habitual grocery, ~15M+ Instacart+ members), though brand advertising contracts are short (<1 year, no long-term commitments) — lower contractual visibility than SaaS, offset by CPG budget stickiness.

Outlook for products/services; how big is the market? Growing. US online grocery ~19% penetrated and growing ~30%; retail media ~$71B (2026, +18%). Fact. Domestic + Canada today (~95% of revenue); early international via enterprise (Costco Spain/France, Instaleap).

Business Quality & Competitive Moat

Industry getting more or less competitive? More competitive in delivery (DoorDash/Uber grocery push, Walmart/Amazon subsidizing), driving defensive price/fee cuts. Retail media is a good pool but ~88% owned by Amazon+Walmart.

How profitable is the business (ROIC/ROE)? ROIC ~14%, ROE ~18–19%, ROA ~11.5%, FCF margin ~24% (Fact, ROIC.ai / 10-K). Respectable for an asset-light platform; ROE flatters on buyback-shrunk equity — use ROIC/FCF.

How profitable is the industry; barriers to entry? Bifurcated: delivery-logistics is low-barrier/low-margin (moatless); enterprise/retail-media has real barriers (data scale, integration switching costs) but is subscale vs the giants.

Can it be easily understood? Yes — a marketplace + advertising + enterprise-tech model. The one subtlety is GTV vs (net) revenue: revenue is only ~10% of GTV because Instacart is an agent, not a grocery reseller.

Undermined by foreign low-cost labor? No — labor is domestic gig shoppers (a regulatory, not offshoring, risk).

Do brands matter? The Instacart consumer brand has modest pull (undercut by low switching costs); the more valuable “brand” is the enterprise trust + data with retailers and CPGs.

Nature of competition; switching costs? Consumer switching costs low (10-K); enterprise switching costs high (deep integrations, “land-and-expand”). This asymmetry is the whole moat story.

Financial Condition & Balance Sheet

Assets not fully on the balance sheet? The 1.6-billion-order proprietary dataset and the retailer/brand relationships are the key off-balance-sheet economic assets underpinning the ad moat.

Off-balance-sheet liabilities? Minimal — small operating/capital leases; gig-worker misclassification is a contingent legal exposure (San Diego $46.5M settled 2024; Prop 22 upheld 2024).

How conservative is the accounting? Reasonable, with the standard caveat that adjusted EBITDA adds back ~$352M of real SBC — track GAAP and cash, not only adjusted. Net (agency) revenue recognition is appropriately conservative.

How CapEx-hungry? Very light — capex ~$61M = 1.6% of revenue. Asset-light platform; growth capital is R&D/SBC and tuck-in M&A, not physical assets.

Capital Allocation & Management

How much FCF, and how is it used? ~$910M FCF (FY2025); used almost entirely for buybacks (~$1.35B in FY2025, funded from FCF + net cash) — no dividend. Interpretation: buyback-first, but only ~46% net-accretive after SBC.

Significant acquisitions? Only small B2B tech tuck-ins (Caper ~$350M/2021; Wynshop ~$106M/2025; Instaleap 2026). Disciplined; no scale-goodwill risk.

Buying back shares / issuing to insiders? Both — ~$2.76B gross buyback over two years, offset by ~$2.8B/yr-equivalent SBC issuance; net ~36M-share reduction over two years. ~$507M of the buyback was privately negotiated insider-exit repurchases.

Compensation policy / motivations? Historically weak pay-for-performance (no performance metrics for NEOs in 2024-25; ~55% say-on-pay in 2025), being fixed under investor pressure (relative-TSR PSUs, boxcar vesting for 2026). Single class of stock, one vote/share — a governance positive. CEO Chris Rogers (built the ads/enterprise engine); founder-CEO Fidji Simo departed for OpenAI (2025).

Valuation & Market Data

ADR / MLP / K-1? No — US common stock, single class, one vote/share. Not an ADR, MLP, or K-1 issuer.

Dividend policy? None, and none intended for the foreseeable future.

How profitable; is net income diverging from cash from operations? GAAP net income $447M vs OCF $971M — cash exceeds net income (non-cash SBC, working-capital timing), a favorable (not red-flag) divergence. GAAP EPS ~$1.57 diluted.

Risks & Downside

What would cause the stock to decline? An advertising-rate disappointment (rate stays ~2.9%), GTV deceleration toward mid-single-digits, deeper margin-eroding price cuts, evidence of agentic-commerce disintermediation, or a large retailer defection — any of which compresses the ~16x GAAP-EBITDA multiple.

Catastrophic-loss risk? Low. Net cash, ~$0.9B FCF, no leverage, asset-light, defensive category — a solvency event is very unlikely. The realistic downside is a valuation de-rate, not a wipeout.

Chance of total loss? Remote — the balance sheet and cash generation preclude it barring an extreme, unforeseeable structural shock.

Recent News & Events

Has the business environment changed recently? Yes — (1) CEO transition (Simo → Rogers, effective Aug-2025; Simo to OpenAI); (2) competitive escalation (DoorDash/Uber grocery push, Kroger exclusivity ended, defensive price/fee cuts); (3) AI/agentic pivot (first app in ChatGPT with Instant Checkout, Dec-2025; Claude integration; Cart Assistant); (4) resilient Q1’26 beat (record GTV $10.29B, adjusted EBITDA +23%).

Significant acquisitions? Wynshop (~$106M, 2025); Instaleap (2026, international fulfillment).

Change in accounting policies? None material.

Recent changes — new markets, facilities, management? New CFO (Emily Reuter, 2024); early international enterprise expansion (Costco Spain/France, Instaleap); new AI Solutions/Cart Assistant; ~$60M regulatory settlements paid Q1’26; a non-recurring Canada DST-repeal benefit.

APPENDIX B — Source Appendix

Maplebear Inc. (NASDAQ: CART) — dba Instacart | Report date: 2026-07-10

All non-obvious facts in the memo trace to a source below. Primary sources (SEC filings, transcripts) take precedence over secondary; third-party aggregated data (ROIC.ai, AZI, FactorsToday, eMarketer) is reconciled to filings where possible. Access date for web/data-feed sources: 2026-07-10 unless noted.

Primary — SEC Filings (Maplebear Inc., CIK 0001579091)

Primary — Transcript

• Instacart Q1’26 earnings call, 2026-05-06/07 (Chris Rogers, CEO; Emily Reuter, CFO). Via ROIC.ai transcript. Used for: forward strategy, advertising diversification and 4–5%-of-GTV target, Cart Assistant (~25% of US customers), 380 storefronts, 310 Carrot Ads partners, 9,000+ brands, Caper 100+ cities, Instaleap, agentic commerce (ChatGPT/Claude), $10-basket lapping, order/AOV dynamics, Q2’26 guidance, $500M revolver, $60M regulatory settlement, Canada DST repeal.
• Instacart Q4’23 earnings call, 2024-02-13 — first full-year adjusted-EBITDA profitability (price-action event #2).

Quantitative Data Feeds (third-party; reconciled to filings)

• ROIC.ai MCP — income statement, balance sheet, cash flow, profitability ratios (ROIC 14.0%, ROE, margins), enterprise value, valuation multiples. Pulled 2026-07-10.
• AZI price history (azitrading.com CSV) — 5-year OHLCV, EMAs, beta/alpha; the price-action Event Map. Pulled 2026-07-10.
• AZI valuation_index — own-history percentile ranks: P/E 26.9x (72.9 pctile), P/B 4.72x (97.5 pctile — flagged as a buyback artifact), P/S 3.37x (62.7 pctile), composite 77.7 pctile. Pulled 2026-07-09/10.
• FactorsToday (factorstoday.com/api) — stock loadings (Market 0.77, Food & Beverage 0.54, Consumer Staples 0.46, Online Retail 0.33; R² ~17%), leaderboard (y1 flat, −36% drawdown; m6 +16.8%; beta 0.77), specific vol ~41%, related stocks (DPZ, SFM, VTEX, TAP, CPRT). Pulled 2026-07-10.

Secondary — Industry & News

• Brick Meets Click / Mercatus — US eGrocery penetration ~19% at YE2025, monthly ~$12.3–12.5B, +29–31% YoY, delivery ~45% of eGrocery. 2025-2026.
• eMarketer — US retail media $60.3B (2025) → $71.1B (2026, +17.8%); Amazon Ads 79.7%, Walmart Connect 8.0%, Target Roundel 1.5%; Instacart & DoorDash ~$1B ads each; Uber Ads ~$1.5B (+60%). 2025-2026.
• Third-party online-grocery share — Walmart 25.7%, Amazon 22.0%, Instacart 21.6% (2025).
• CNBC / Bloomberg — CEO transition (Rogers for Simo→OpenAI), 2025-05-28; FTC pricing-tool probe, 2025-12-17.
• PR Newswire — Uber Eats partnership (2024-05-07); Instacart app in ChatGPT / Instant Checkout (Dec-2025).
• Grocery Dive / Supermarket News / Progressive Grocer / Investing.com — quarterly prints, Kroger exclusivity end, San Diego $46.5M worker-misclassification settlement (2024), Prop 22 upheld (Jul-2024), Wells Fargo price/fee-cut channel note.
• TechCrunch / Grocery Dive — M&A history (Caper AI ~$350M/2021; FoodStorm, Eversight, Rosie; Wynshop 2025; Instaleap 2026).