Selling More Does Not Fix Selling at a Loss: The Argument for Margin Before Scale

The Editorial Desk's Position

Redseer's September 2026 forecast, covered in our Retail package, said quick commerce averaged more than 60 million monthly transacting users in the first half of 2026 and a market of about US$9 billion in that half, doubling for the third straight year. The format has changed how Indians shop. We are not sceptics of that.

We are sceptics of a particular inference: that a fast-growing business with large losses will grow into profit simply by growing. The evidence we have seen does not show that, and we think founders, investors and readers should ask a more specific question.

What the Filings Showed

Inc42, reading Zepto's FY26 filings in August 2026, reported operating revenue of Rs 22,623.58 crore and a net loss after exceptional items of Rs 5,905.19 crore, compared with Rs 4,699.71 crore in FY25 and Rs 1,214.79 crore in FY24. Total expenditure was Rs 29,026.75 crore. Sales of traded goods were Rs 17,587.92 crore, 77.7% of operating revenue, and services revenue Rs 5,022.52 crore, 22.2%, up from Rs 547.61 crore in FY24. The cost of goods purchased was Rs 18,484.98 crore. Inc42's sharpest observation was that after adjusting for inventory change, the net cost of goods exceeded traded-goods sales by about Rs 610 crore, in its words: "The company is losing money on goods sold before even counting delivery rider pay, rent, or app costs."

That is Inc42's calculation from the filings, not a company statement, and we have not seen the company's reply.

“A business that loses money on every order does not fix it by taking more orders.”

The Argument: Margin Before Scale

A business whose gross margin is negative loses more as it sells more, unless something else changes. Possible fixes include higher prices, a richer product mix, bigger baskets, lower supplier costs, or income from services such as advertising and fees. The filings suggest the company is pursuing the last of these: services revenue rose from Rs 547.61 crore in FY24 to Rs 5,022.52 crore in FY26, with advertising at Rs 1,635.73 crore. But Inc42 estimated net advertising contribution, ad revenue less ad spend, at about Rs 247 crore, which is small beside a Rs 5,905 crore loss.

The proposed Code on Social Security levy of 1% to 2% of turnover for gig-worker benefits, capped at 5% of payments to workers, which Inc42 estimated at Rs 226 to Rs 452 crore at the lower end on FY26 revenue (Inc42's own calculation), adds to the cost side.

Our position is that unit economics should be tested before scale, not after. In the Scaling package we made that point with a hypothetical service business, where a small change in one input turns a loss-making unit profitable. The method is to find that input first.

What Scale Can Do & The Strongest Counter-Argument

What scale can do: We do not say scale never helps. Density lowers delivery cost per order, bargaining power lowers supplier prices, and a larger audience raises advertising value. These are real effects. What we say is that they need to be shown in the numbers, store by store or cohort by cohort, before they can be assumed.

The strongest counter-argument: The best reply is that building a network is capital-intensive and loss-making by nature, as it was for other platforms elsewhere; that early gross margin is a poor guide because the company is subsidising to win customers; and that a ratio calculated from filings, adjusted for inventory, is a snapshot of a rapidly changing business. Redseer's description of the market as growing from a large base supports the view that the opportunity is real. Our Scaling package noted that Blinkit guided to 3,000 stores by March 2027, which is a statement of management confidence in store economics, though a guidance and not a result.

We accept that a negative gross margin in one year does not prove that the model is broken. We argue that it should be treated as a hypothesis the business must demonstrate it can reverse.

What Would Change Our Mind & Lessons for Founders

What would change our mind: If subsequent filings showed gross margin turning positive on a like-for-like basis as the business grows, or if cohort data showed order margins improving with age, we would say the model is working. If the company's listing process produces audited information that contradicts Inc42's calculation, we would update the piece.

What it means for founders: Ask of any growth plan: what is the margin on the marginal order, and how does it change as we double? If you cannot answer, find out before raising the next round.