Zepto raised at roughly $5 billion in late 2024, and investors keep buying the story. The pitch is clean: Indians now expect groceries in ten minutes, the category is vast, and whoever wins dark-store density wins the decade. All of that could be true and the investment could still be mediocre. This is a case about the difference between a real business and a reasonably priced one.
What they actually do.
Zepto operates a network of small, densely placed warehouses — "dark stores" — across Indian cities. Each stocks roughly 3,000 SKUs of fast-moving groceries, essentials, and increasingly general merchandise. Orders come through the app, get picked in under two minutes, and reach the customer in ten. The entire model lives or dies on store density: thinner networks mean longer trips, longer trips mean worse economics.
The category is a three-horse race: Blinkit (Zomato), Instamart (Swiggy), and Zepto. Nobody is profitable at a net level yet. Zepto claims positive contribution margin at the store level and a path to overall profitability — a claim worth examining.
The numbers, one line at a time.
Every quick-commerce order carries roughly the same structure. A customer spends an average of about ₹450. Gross margin on groceries runs 18–22%, so contribution before delivery cost sits around ₹85–100.
Then comes the cost stack. Rider economics: a rider doing four orders per hour at ₹25 per drop costs ₹25 per order at full utilization — more when demand softens. Add dark-store opex (rent, electricity, pickers, breakage) of ~₹15 per order at maturity, and store-level contribution lands at ₹45–60 per order, before central marketing and HQ overhead.
| Line item | Per order (₹) | % of AOV |
|---|---|---|
| Average order value | 450 | 100% |
| Cost of goods | (360) | (80%) |
| Gross contribution | 90 | 20% |
| Delivery rider cost | (25) | (5.5%) |
| Dark store opex | (15) | (3.3%) |
| Store-level contribution | 50 | 11.1% |
| Marketing & CAC (allocated) | (30) | (6.7%) |
| HQ, tech, finance overhead | (20) | (4.4%) |
| Net contribution | ~0 | 0.0% |
Read the table honestly. Even on friendly assumptions, net margin is approximately zero at steady state. Every rupee of profit requires either (a) raising AOV through higher-margin categories — beauty, electronics, private label — or (b) reducing per-order delivery cost through density and automation. Both are real levers. Neither is guaranteed.
The business is real. The multiple is the question.
Orders per dark store per day.
Every quick-commerce operator's path to profitability runs through one number: daily orders per dark store. At 800, the model works. At 500, it bleeds. At 1,200, it's a generational business. The reason is fixed cost absorption — rent, picker wages, and store manager salary are roughly constant whether you do 300 or 1,300 orders.
Zepto's reported density improved from ~450 orders/store in 2022 to near 900 in its best cohorts by 2024. That's real operating leverage. It is also the only thing keeping the store-level P&L above water. If density flattens for four quarters, the entire profitability thesis gets repriced.
Both cases, laid out flat.
Bull Case
- Category is structurally winner-take-most — density begets density, economics reinforce.
- AOV expansion into beauty, apparel, private label adds 400–600 bps gross margin.
- Automation (pick-and-pack robotics) cuts 20–30% of variable cost by 2028.
- Ad monetization becomes a third revenue line at 90%+ margin.
- Indian consumer moves to higher frequency of small-basket purchases, growing TAM 3x.
Bear Case
- Contribution margin plateaus at 10–12% — not enough to cover HQ at any realistic scale.
- Three-player war keeps CAC elevated indefinitely; nobody gets to rational pricing.
- Dark-store real estate inflates as the model expands; unit economics degrade.
- Blinkit's Zomato cross-subsidy makes Zepto's standalone math structurally worse.
- Regulatory tightening on gig-worker classification adds 15–25% to delivery cost overnight.
The call.
A real business — with a priced-in path to profitability.
Zepto is not a scam and it is not a bad business. It is a plausibly good business priced as if the outcome is certain. The $5B valuation requires the bull case to play out cleanly — order density continues expanding, AOV shifts toward higher-margin categories, and the competitive war calms enough to let pricing rise 4–6%.
Any one of those breaking delays profitability by 18–24 months. Two of them breaking makes the current valuation indefensible. If you are evaluating whether to invest as an employee, angel, or secondary buyer: the question is not whether Zepto will exist. It will. The question is whether the next dollar of capital earns a return above 15% from a $5B entry price. The honest answer is: only if everything goes right.