Business

Claude View

Know the Business — Eternal Ltd (ETERNAL)

Eternal is no longer "Zomato". It is a conglomerate-in-the-making where a mature food-delivery cash engine (India duopoly with Swiggy, ~4-5% adjusted EBITDA margin on NOV, compounding ~20% a year) is being used to fund a land grab in quick commerce (Blinkit, 2,027 dark stores, 121% YoY NOV growth in Q3FY26, just reached adjusted EBITDA breakeven). The market is paying ₹1,040x trailing earnings not for food delivery; it is paying for Blinkit becoming the Meituan of India. Everything else — District, Hyperpure, Bistro — is optionality stapled onto the same consumer graph.

Q3FY26 changed the story in two ways: Blinkit flipped to positive adjusted EBITDA (+₹4 Cr) a quarter earlier than consensus; and founder-CEO Deepinder Goyal stepped down effective 1 Feb 2026, handing operations to Blinkit CEO Albinder Dhindsa. The stock now trades at ~₹249 versus a 52-week high of ₹368 — the market is openly debating whether a Blinkit-led Eternal can sustain its growth without discounting its way into the same margin trap that collapsed China's Meituan.

How This Business Actually Works

Eternal is four separately-managed businesses sitting on one consumer login. Food delivery and Blinkit generate 85%+ of adjusted revenue. Both are platforms, but the unit economics are fundamentally different — one is an asset-light marketplace taking a clip on restaurant GOV, the other is a real-estate-plus-inventory business where the company owns the dark store, pays the rent, buys the stock and rents the rider.

The food delivery engine. Zomato hosts 297k monthly active restaurants and 473k monthly active delivery riders across 800+ cities. The platform does not own inventory or riders. It collects a commission (~18-22% of order value depending on restaurant tier) plus ad revenue, plus a platform fee per order (₹10-15, raised multiple times in FY24/FY25), plus a customer delivery charge. Adjusted revenue take-rate was 28.7% of NOV in FY25. Strip out last-mile rider payout and you are left with contribution margin of 9.4% of NOV (₹3,092 Cr). Fixed overhead — tech, corporate, support — gets covered, leaving a 4.6% adjusted EBITDA margin on NOV in FY25, rising to an all-time-high 5.4% in Q3FY26. Management guides 5-6% as the steady state. The incremental-profit equation is brutally simple: each order adds maybe ₹20-25 of contribution; cost per delivery has been falling through batching and EV penetration (37k EV riders; last-mile emissions down 10.5% vs FY22). Growth is a straight function of order growth × AOV growth, and in FY25 that was 13% × 6% = 20% NOV growth. Going forward, management targets 20%+ NOV growth powered by rising MTC (20.6m, +12% YoY) and higher average frequency per user.

The Blinkit engine. Quick commerce is a completely different animal. The product is "a bottle of shampoo in 10 minutes at MRP", and the business is a network of 2,027 dark stores (as of Q3FY26, up from 1,301 in Mar-25, headed to 3,000 by Mar-27) each serving a 2-3 km radius with ~5,000-7,000 SKUs and 2-3 riders constantly in motion. Starting Q1FY26 Blinkit shifted from a marketplace (taking commission on seller-owned inventory) to owning ~90% of inventory itself — that is the main reason reported "revenue" exploded 756% YoY in Q2 and 848% YoY in Q3FY26 without any real change in economics (GOV growth was 121% YoY). The unit economics that matter are:

A mature Blinkit dark store in Gurgaon or Noida does ~₹12-15 lakh of daily GOV and runs at ~5% adjusted EBITDA margin; the blended network did ₹963k GOV/store/day in FY25 pulled down by 775 newly opened stores. Mature Delhi NCR operates at ~3.5%, Gurgaon + Noida at ~5% — management's entire case rests on the rest of India following that same maturation curve. Capex per store is rising (bigger stores, more SKUs, better picking tech) but Akshant Goyal (CFO) guides 40%+ ROCE on incremental capex + NWC in steady state. The true cost stack per order is roughly ₹50 in dark-store rent & staffing + ₹35 last-mile delivery + ₹5-8 customer-acquisition subsidy + wastage and packaging — against a net AOV of ₹547 and a take-rate in the low double digits. Profit is a fight over 3-4 percentage points.

Bargaining power is asymmetric. Restaurants on Zomato have almost no alternative — a chef skipping the platform loses 40-60% of their delivery sales overnight. Delivery riders are the weak link: monthly active rider base grew 18% YoY but churn is reportedly 8-10% a month, and the Nov-2025 notification of the Code on Social Security creates a new cost overhang (Eternal spent ₹100+ Cr on rider insurance in CY2025; the code may push this by a similar order of magnitude). Consumers are promiscuous — Zomato Gold (12.1m members) is the retention fence. In Blinkit, the vendor relationship inverts: Blinkit is the customer (it owns inventory), so FMCG brands negotiate hard for shelf placement and advertising; consumers are even more promiscuous because everyone is selling Maggi at MRP. The moat in quick commerce is density of dark stores per sq km in a specific city, not brand — which is why Blinkit, Instamart, and Zepto keep opening stores on top of each other in South Delhi and Bangalore.

The Playing Field

Screener lists six peers, but only one of them — Swiggy — actually competes with Eternal for the same rupee. Nykaa, Meesho, FirstCry, CarTrade and Urban Company are adjacent consumer-internet names. The real competitor set is a mix of listed companies, privately-held scale-ups (Zepto, Meesho pre-IPO) and deep-pocketed mega-platforms (Flipkart Minutes, Amazon Now, JioMart).

Eternal is the largest consumer-internet company in India by market cap after the private tech giants. It trades at 7.8x book vs Nykaa's 54x — a misleading comparison because Eternal issued ₹8,500 Cr of QIP stock in Nov-24 (cash sits on the balance sheet, inflating book). Head-to-head, only Swiggy operates in both of Eternal's core markets, and Eternal outscores it on every operational metric that matters.

The gap is not cosmetic. In food delivery, Zomato's 20% NOV growth in FY25 outpaced Swiggy's ~17%, and the margin delta (4.6% vs 1.6% of NOV) means Zomato extracts roughly 3x the operating cash per order. In quick commerce, Blinkit has close to 2x the store count of Instamart and produces positive adjusted EBITDA while Instamart still loses ~15-18% of NOV. Bernstein's call that the "leaders" (Blinkit, Instamart, Zepto) are pulling away from "challengers" (Flipkart, Amazon, JioMart, BigBasket) is broadly right, but inside the leader set Blinkit is alone in having ignited the operating-leverage flywheel. That is what ~₹2.4 lakh crore of market cap is expressing.

The chart is directional, not derived. Scale proxies for NOV/GMV; profit execution proxies for proximity to positive unit economics. Eternal sits alone in the top-right because it has both scale and actual EBITDA positive economics across two businesses. Swiggy and Zepto are sub-scale-to-Blinkit in stores and further from unit-economic parity. Nykaa is profitable but in a much smaller addressable pool. The quadrant matters because quick commerce is converging toward a 3-player equilibrium (Bernstein, Kotak), and the player that enters steady state first can self-fund while rivals raise equity.

Is This Business Cyclical?

Less than people assume, for different reasons across segments.

Demand cycle. Indian urban food delivery has behaved like a staple since 2023 — even during the weak discretionary quarters of 2024 (when Swiggy, DMart and HUL all flagged "urban slowdown"), Zomato food-delivery NOV dropped from 27% YoY growth in FY24 to 20% in FY25 and then 13.1% in Q1FY26 before recovering to 16.6% in Q3FY26. A slowdown, not a crash. Quick commerce is stickier than it looks — the penetration story is still early (10.2m MTC in FY25 vs 100m+ e-commerce users) — but the basket skews discretionary (apparel, electronics, personal care), making it more exposed to consumer sentiment than a grocery store like DMart. Management's explicit signal is that demand is sensitive to competitive pricing: heavy discounting by Amazon / Zepto stimulates demand in short bursts that is "less durable".

Monsoon and festive seasonality. Both businesses spike in Oct-Dec (Diwali + IPL + cricket). Q2FY26 pulled festive demand forward and Q3FY26 softened consequently. Monsoon (Jun-Sep) is generally positive for food delivery (people order in), negative for quick commerce (perishables / compliance). Delhi NCR pollution and the GRAP regulations cost Blinkit ~70 new-store openings in Q3FY26 — an underappreciated seasonal risk.

Cap-market cycle — the one that actually matters. Both Zomato (Jul-2021) and Swiggy (Nov-2024) IPO'd into generous new-age-tech valuations. Eternal raised ₹8,500 Cr QIP in Nov-24; Zepto raised $1.3 bn; Swiggy has its own cash cushion. That collective capital is what funded 2025's discount war. If the IPO/funding window stays open, discounting continues and margin expansion is delayed. If it closes, Blinkit's scale advantage becomes mechanical: it has ₹17,820 Cr of cash (Dec-25), positive EBITDA, and can simply outlast sub-scale rivals. That is why the stock is a leveraged bet on Indian consumer-tech capital discipline.

The cycle inside the cycle is competitive intensity — growth decelerates because leaders are becoming more disciplined, not because the market is saturating.

The Metrics That Actually Matter

Forget consolidated operating margin. It is a meaningless blended number while Blinkit is scaling. The five metrics below are what the management team, the buy-side and the bear case all fight over.

MTC is the honest demand metric. Food delivery MTC growth of 12% YoY in FY25 and 21% YoY in Q3FY26 is the cleanest read of actual customers, stripped of AOV games and platform-fee changes. The 2x growth in Blinkit MTC (from 10.2m to 23.6m over four quarters) is the story that matters — it is the closest thing to proof that quick commerce is an actual consumer behavior change, not a venture-funded mirage.

Contribution margin, not adjusted EBITDA. Adjusted EBITDA is noise-adjusted for growth investments. Contribution margin — what is left after variable costs including rider payout, dark-store cost and discounts — is the true test of unit economics. Blinkit went from 2.5% to 4.3% to 4.9% of NOV over six quarters. That is what tells you the model works. Food delivery's 9.4% contribution margin is already mature.

GOV per dark store per day. The denominator of quick-commerce economics. Blinkit's blended figure is ~₹12.7 lakh per day; Instamart's is ~₹8.5 lakh. The gap is why Blinkit turned EBITDA-positive and Instamart has not. Watch for GOV/store/day to decelerate as Blinkit adds 1,000 new stores in FY27 — it will pull the blended number down even if same-store sales grow.

Delivery cost per order. Not disclosed directly, but derivable: food-delivery "delivery and related charges" of ₹5,728 Cr in FY25 ÷ (853m food orders + 424m quick-commerce orders) ≈ ₹45 per order on a blended basis. This number falls with batching (Blinkit's 2-3 orders per rider-trip vs food delivery's 1), EV penetration, and AI route optimization — but rises with the social security code.

Blinkit steady-state NOV times target margin. The bull math: if Blinkit reaches ₹1.5 lakh crore NOV by FY30 at 5.5% adjusted EBITDA margin, that is ₹8,250 Cr of EBITDA from just Blinkit, versus current consolidated adjusted EBITDA of ₹1,079 Cr. That is the implicit number the ₹2.4 lakh crore market cap is betting on.

What I'd Tell a Young Analyst

First, forget the food-delivery business for purposes of the stock's directionality. It is a beautiful, cash-generative, 20%-growing duopoly with a real moat — and it is roughly 40% of intrinsic value. The stock will move on Blinkit, period. Valuation work should separately DCF Zomato food delivery at 25-30x steady-state EBITDA (it is a Nifty-quality compounder) and SOTP Blinkit as a function of terminal NOV × 5-6% margin × multiple. Everything else (District, Hyperpure, Bistro) goes into option value.

Second, the single question that dominates the thesis for the next 18 months is: can Blinkit compound NOV at 80%+ without sacrificing contribution margin? Management hit breakeven in Q3FY26 by being disciplined on customer acquisition; the bear case is that Amazon Now, Flipkart Minutes, and a recapitalized Zepto force Blinkit back into discounting. Track contribution margin quarterly — it will crack before adjusted EBITDA does. If it falls below 3.5% of NOV for two consecutive quarters with revenue growth still decelerating, the thesis is broken. 5%+ sustained means the flywheel is real.

Third, the CEO transition is a bigger deal than the market has digested. Deepinder Goyal built the company, survived two near-death moments (FY19 and FY22), pulled off the Blinkit acquisition at the exact trough of quick commerce, and returned 5x from IPO. He is staying as vice-chairman but the operator is now Albinder Dhindsa — who runs the business that matters, so the continuity risk is low but Goyal was the capital-allocation brain. The next big call — doubling down to 4,000 stores vs banking margins at 3,000 — falls to Dhindsa + Akshant Goyal (CFO). Watch the Q1FY27 (Jul-26) letter for tone on store count and discounting.

Fourth, the things the market is probably under-weighting: (a) the Code on Social Security for gig workers — it is coming, it will cost something, and while management is sanguine, a 100 bps hit to food-delivery margin is plausible; (b) GST litigation on delivery charges (₹441 Cr in active disputes, contingent) and the broader regulatory drag on platform fee; (c) the fact that Blinkit's inventory shift now means real working capital + inventory risk sitting on Eternal's balance sheet — FY26 will be the first year that shows up in debtor/inventory days and cash conversion (already deteriorating: debtor days 35 vs 24 in FY24).

Fifth, the things the market is probably over-weighting: generic "competition" fear. India has run three cash-burn cycles in food delivery and three players emerged (Zomato, Swiggy, one more). Quick commerce will likely settle at 3-4 viable players. Blinkit having 2x Instamart's stores, breakeven economics, and the lowest cost of customer acquisition is the winning hand. The 52-week drawdown from ₹368 to ₹249 is mostly competitive-intensity fear, not a broken business.

Final rule: this is not a P/E stock. P/E of 1,040x is meaningless when 80% of future profit pool is in a segment that just turned EBITDA-positive this quarter. Value it on NOV × margin × multiple by segment, not on trailing earnings. The right valuation question is what multiple of FY28 EBITDA will the market pay when Blinkit is at 4% margin on ₹70,000 Cr NOV? — and that answer, not the current P/E, is what determines the return.