The Execution Premium: Why India’s Best Ideas Now Win or  Lose on the Ground

On 28 April 2026, Eternal Limited reported a number that settled an argument the Indian startup ecosystem had been having for four years. Blinkit, its quick commerce arm, had turned adjusted EBITDA positive for the first time on a full-quarter basis, posting a profit of ₹37 crore on net order value of ₹14,386 crore in the March quarter, against a loss of roughly ₹178 crore a year earlier. Eternal’s consolidated net profit rose to ₹174 crore from ₹39 crore.

Almost nothing about the underlying idea was original. Ten-minute delivery had already been attempted and buried elsewhere. Getir and Gorillas raised billions of dollars in Europe and collapsed. The concept travelled across borders effortlessly, as concepts do. What refused to travel was the ability to actually run it.

That is the uncomfortable centre of this theme. Ideas are abundant, cheap and increasingly commoditized. Execution is scarce, expensive and stubbornly local.

The scoreboard nobody puts on a slide

India’s quick commerce market is now worth roughly ₹1.08 lakh crore within an ₹8 lakh crore digital commerce market, growing at about 40 percent year on year, more than twice the pace of digital commerce overall, according to an Equirus report released in July 2026. The visible driver is not strategic brilliance. It is concrete and shelving. The combined dark store networks of Blinkit, Instamart and Zepto expanded from 3,405 locations in May 2025 to 5,026 in May 2026, an increase of nearly 48 percent in twelve months.

Every one of those stores represents a stack of operating decisions: which 3,000 SKUs to stock in that specific pin code, how to roster shifts around a 7 p.m. demand spike, how quickly to replenish before a Sunday washout, how to keep pick-and-pack error rates low enough that refunds do not eat the contribution margin. None of that is transferable through a strategy deck. It is earned store by store, quarter by quarter, and it is precisely why Blinkit’s per-store economics, rather than its ten-minute promise, became the moat.

The counter-example sits in the same market

Contrast this with the Open Network for Digital Commerce. Arguably ONDC had the better idea: an open, interoperable protocol to break platform monopolies and give kirana stores the reach that Amazon and Flipkart reserve for scale sellers. It scaled from a five-city pilot to over 600 cities. Then retail momentum stalled. ONDC’s retail orders fell from 6.5 million in October 2024 to 4.6 million in February 2025, and retail’s share of network transactions dropped from 47 percent to 29 percent.

What grew instead was mobility, which rose from 3.6 million orders in April 2024 to 9.4 million in April 2025. The reason is instructive. Booking an auto involves no inventory, no returns, no refunds and no cold chain. Retail involves all four. The idea did not fail. The operational load it placed on thousands of small, under-resourced participants was simply heavier than the network could carry.

Same country, same digital rails, same consumers. One built an execution machine. The other built an elegant architecture and underestimated the plumbing.

The gap is structural, not Indian

This is not a local failing. Harvard Business Review research widely cited since 2016 estimates that 67 percent of well-formulated strategies fail because of poor execution. Bridges Business Consultancy found that 48 percent of organisations miss at least half their strategic targets, and that only 7 percent of leaders consider their organisation excellent at implementation. Its earlier surveys found that 85 percent of leadership teams spent under an hour a month discussing strategy execution.

The pattern is consistent. Organisations invest heavily in deciding what to do and almost nothing in the far duller discipline of ensuring it happens the same way on the 400th day as on the first.

What this should change

Three practical shifts follow.

First, treat execution capacity as a real constraint on strategy, not an afterthought. A plan that assumes flawless frontline behaviour is not a plan, it is a wish.

Second, respect the human system underneath. India’s dark store sector directly employs an estimated 40,000 to 80,000 workers but generates an annual hiring demand of 110,000 to 220,000 because attrition is brutal. Execution quality lives with people who are trained fast, paid modestly and replaced often. Every rupee of operating leverage is mediated by them.

Third, measure consistency, not just velocity. Blinkit’s turnaround came from adding stores while narrowing losses per store, a trade-off that only shows up if you are tracking both.