India’s e-com boom puts fulfilment and returns under the microscope

The true economics of fulfilment: A parcel arriving at a doorstep is the final visible moment of an e-commerce transaction. But, behind it sits a far more complex chain of inventory, warehouses, transport networks, delivery attempts and, increasingly, returns. In fact, the journey begins well before a customer clicks “buy”. Inventory must be positioned close enough to demand, picked and packed efficiently, routed through the appropriate delivery network and delivered within the promised window. A wrong address, an unavailable customer or a failed delivery attempt can set off another cycle of transportation, handling and cost.

And when a parcel comes back, the journey is not over yet. The product has to be received, scanned, inspected and graded before it can return to saleable inventory, undergo refurbishment, be liquidated or be written off.

That complexity is growing with the market. The Google-Deloitte The $250 Billion Commerce Frontier report projects India’s e-commerce market to reach $250 billion by 2030, from $90 billion today, with 150 million additional shoppers expected to enter the digital economy and individual e-commerce spending projected to double.

The growth is also spreading beyond the metros. More than 60% of e-commerce transactions are associated with Tier-II and Tier-III cities, according to Deloitte-FICCI, while Shiprocket says these markets accounted for 61% of its shipments in FY26. So, the question for logistics is no longer simply how to deliver more. It is how to make every movement count.

When a delivery does not reach the customer
For an e-commerce business, a failed delivery can cost far more than the original transport charge. The seller has already paid to move the parcel from the fulfilment centre towards the customer. If delivery fails, another attempt may be required; if it ultimately returns, reverse transportation adds to the bill.

The cost continues at the warehouse, where returned products require handling, inspection and processing before they can potentially be sold again. The longer the cycle, the longer the seller’s money remains tied up in unavailable inventory.

Indicative fulfilment economics put forward shipping at ₹50–80 per order and return-to-origin (RTO) reverse logistics at ₹40–60 per failed delivery. Repackaging labour can add ₹15–25 per returned unit, while working capital can remain tied up for 7–14 days. Product damage is estimated at 3–8% of returned units.

At a RTO rate of 20–25%, the indicative blended cost per delivered order can reach ₹85–110. Meanwhile, DPIIT-NCAER estimates India’s aggregate logistics cost at ₹24.01 lakh crore in 2023-24, or 7.97% of GDP.

The first return is often the most expensive one
The most efficient return is one that never happens. Ajay Rao, Founder & CEO, Emiza, says the company sees a blended return rate of approximately 15% across its network, which has remained broadly stable over the past few years. But the number varies significantly by category. In fact, fashion and apparel record return rates of around 35–40% across Emiza's operations, driven by multiple size choices, fit, fabric expectations and customers changing their purchase decision after ordering.

Not all returns create the same operational workload. Courier returns or RTO shipments are relatively straightforward when the package remains sealed and has not been handed over to the customer. It can be returned to the warehouse, checked and, where appropriate, put back into stock. However, customer-initiated returns are more complex. Products may have been opened or used, customers may expect immediate refunds, and logistics providers have to guard against product swaps, incorrect returns and other forms of return-related fraud.

At Emiza, field teams conduct detailed quality checks during pickup, while returned products are subsequently inspected and graded at the warehouse. The target is to return inventory to a saleable state within 48 hours. Products with minor wear, damaged packaging or missing tags may undergo cleaning, repackaging or other corrective measures, while severely damaged products may be written off or discarded. The faster that decision is made, the shorter the inventory recovery cycle.

Preventing the return before it starts
The stronger focus across logistics operators is shifting from managing returns to preventing avoidable failed deliveries. Blue Dart has identified wrong addresses, missed re-attempts and incorrectly dispatched products among recurring causes.

“Returns move through structured pickup, scanning and tracking workflows on our end, and we support both open and closed return models. Our bigger focus is on preventing the RTO before it happens rather than managing it well after the fact - the most common preventable causes tend to be wrong address at checkout, missed re-attempts on failed deliveries, and wrong product dispatched,” said Sonia Nair, National Head – Customer Service, Blue Dart.

“We work on all three: tighter address verification, proper re-attempt processes, better exception handling. It's a shift we'd like to see more industry-wide; returns don't have to be treated purely as a cost of doing e-commerce,” added Nair.

Emiza uses address-validation application programming interfaces to identify inaccurate or incomplete addresses, while customer and transaction data supports artificial intelligence-based risk scoring. Repeated non-acceptance or suspicious behaviour can trigger one-time-password confirmation or prepaid payment. When a shipment is already at risk, Emiza's in-house non-delivery report team contacts customers proactively. The company says it successfully recovers approximately 50–60% of such shipments, allowing them to be delivered rather than sent back. That intervention can eliminate an entire reverse-logistics movement.

“Fashion and apparel consistently record the highest return rates, which can reach 35–40%.”
Ajay Rao, Emiza

When a return is unavoidable, speed of recovery becomes critical. Emiza’s 48-hour warehouse target is built around this principle. Every returned unit is inspected and graded: sellable products go back into stock, those needing minor intervention are refurbished, while others are liquidated or written off.

The faster this process moves, the less time inventory remains outside the available stock pool. Returns can also carry costs beyond transport, including multiple delivery attempts, reverse transportation, warehouse handling, inspection, repackaging, refurbishment, idle inventory and cash-flow impact. Preventing avoidable returns also removes the reverse journey, reducing vehicle movement, fuel consumption and additional packaging. Technology is being applied across the order lifecycle. Blue Dart uses AI and predictive analytics for demand forecasting, route optimisation and resource planning, alongside automation in sorting.

“Technology has to solve a real operational challenge before we invest in it. We use AI and predictive analytics for demand forecasting, route optimisation, and resource planning, enabling us to anticipate demand and improve network efficiency,” said Nair. She also added, “When it comes to generative AI, we are taking a deliberate, use-case-led approach, focusing on long-term value rather than rapid deployment.”

How can checkout become a logistics intervention point?
Shiprocket is applying the same principle before an order becomes a shipment. Saahil Goel, Managing Director & CEO, Shiprocket, stated in a report that the e-commerce ecosystem has evolved from simply providing customer access towards integrating technology across discovery, checkout and fulfilment. He said, “At Shiprocket, we have worked with more than 400,000 businesses across India.”

“Over the last decade, millions of entrepreneurs have gained access to customers they could never reach before. A business in Karnal can now sell to Kochi. Homegrown D2C brands can find customers across the country without opening a single physical store. Digital infrastructure has levelled the playing field in remarkable ways. And, the next phase of this journey will be defined by intelligence,” added Goel.

Shiprocket says artificial intelligence can reduce checkout friction, encourage customers to shift from cash-on-delivery to prepaid orders and recover abandoned carts. The potential financial impact can be significant even with small improvements. A 1% conversion improvement for a store generating ₹1 million in monthly sales could add ₹10,000 in monthly revenue without additional traffic spending. A 25% reduction in RTO shipments for a business experiencing ₹30 million in annual failed deliveries could save ₹7.5 million a year.

In one company example, checkout redesign, AI-enabled payment optimisation, and cart recovery increased conversion from 1.1% to 2.4%. RTO fell from 31% to 22%, repeat purchase exceeded 28%, and revenue increased by 47%, without a material increase in advertising spend. These are company-specific examples, not industry-wide benchmarks, but illustrate how logistics and customer experience are becoming connected parts of the same commercial process.

E-commerce growth has also made courier selection more complex. Shiprocket states that there are more than 25 major courier partners in India, each offering different pricing, service levels and delivery performance. Artificial intelligence can dynamically select partners based on location, capacity and expected delivery performance, matching each shipment with an appropriate network rather than relying on one provider for every order. This becomes increasingly relevant as e-commerce expands into smaller cities and towns, where shipment density and delivery economics can differ from metropolitan markets.

Inventory is moving closer to the customer
The geographic expansion of demand is also changing warehouse strategy. Ramnath Subramaniam, Joint Managing Director, TVS Industrial & Logistics Parks, said customers are increasingly positioning inventory closer to emerging demand centres. The company has observed this among major FMCG and e-commerce customers, including Nestlé and Flipkart.

“The expansion of capacity across markets such as Hosur, Chakan, Cuttack, Coimbatore and Siliguri reflects growing need for infrastructure closer to developing consumption and industrial centres.”
Ramnath Subramaniam, TVS ILP

Its expansion markets include Hosur, Chakan, Cuttack, Coimbatore and Siliguri. The logic is straightforward: as demand becomes more distributed, keeping inventory concentrated around a limited number of metropolitan centres can increase delivery distance and cost. The response is a more distributed warehousing network.

TVS ILP is developing Grade-A industrial and logistics parks in emerging markets, focusing on network density across Tier-II and Tier-III locations. This does not make large fulfilment centres redundant; instead, different facilities increasingly serve different roles.

“Quick commerce has redefined customer expectations around convenience and visibility, but we do not believe every shipment will move to a quick-commerce model,” Nair said. She added, “What it has influenced more significantly is how businesses approach inventory placement, fulfilment networks, and delivery speed.”

Conventional fulfilment centres can hold large volumes and serve wider geographies, while smaller last-mile facilities can position selected inventory closer to high-frequency demand. According to Subramaniam, quick commerce requires a more distributed and interconnected infrastructure comprising large warehouses, fulfilment centres and last-mile facilities. The company cited a warehouse requirement involving TCI and Zepto in Coimbatore as an example of changing requirements in emerging markets. As inventory becomes more distributed, real-time inventory visibility and faster replenishment become increasingly important.

The changing fulfilment model is also reshaping logistics real estate. TVS ILP is incorporating automation-ready infrastructure, digital facility management and sustainability features, including its One TVS ILP digital and AI application, EDGE-certified green buildings, solar provisions and rainwater harvesting. TVS ILP plans to invest ₹2,500–2,700 crore over the next three years, expanding its portfolio from approximately 12 million sq ft to 20 million sq ft by 2028, with a significant focus on Tier-II and Tier-III markets. The underlying requirement is flexibility as inventory, technology and customer needs evolve.

Is air cargo the only option for e-com?
E-commerce growth does not always mean every parcel moves by the fastest available mode. For instance, Blue Dart's shipment volumes are split almost equally between air and surface transportation.

“Our bigger focus is on preventing the RTO before it happens rather than managing it well after the fact.”
Sonia Nair, Blue Dart

“Our shipment volumes are split almost equally between air and surface transportation, and in our network, these two modes complement each other exceptionally well. While air enables speed and time-critical deliveries, surface provides scale, reach, and cost efficiency,” Nair said. “Together, they create a balanced and resilient network that allows us to consistently meet diverse customer requirements, with multimodal solutions further enhancing flexibility where needed,” she added.

Surface B2B and e-commerce were major growth drivers for Blue Dart in the past quarter, with air continuing to serve time-critical shipments while surface offers greater cost efficiency. Nair said the two modes can be combined through multimodal solutions, an approach becoming increasingly important as e-commerce expands into smaller markets. The challenge is to match delivery requirements with the most appropriate balance of speed and cost.

Festive demand tests every layer
The festive period puts fulfilment infrastructure under pressure as demand rises across multiple categories simultaneously. Blue Dart says e-commerce accounts for approximately 30% of its revenue and was one of its consistent growth drivers in the first quarter of FY2027. While it does not separately disclose e-commerce shipment volumes, the company moved more than 400 million domestic shipments across its network in the past year.

“Festive season is when e-commerce demand shows up most sharply, and it's shaped how we plan capacity - rather than building around a single spike, we plan for multiple demand peaks through the year and make sure capacity sits in the right markets rather than just adding more of it everywhere,” Nair said.

Nair mentioned that Blue Dart's dedicated air network comprises eight cargo aircraft—six Boeing 757-200 freighters and two Boeing 737-800 freighters—operating 26 flights a day across more than 56 routes, with over 700 tonnes of payload capacity per day. The network is supported by more than 33,000 vehicles, over 60 hubs and gateways, and additional manpower across operational and customer-facing functions.

But, festive planning begins months ahead, with capacity positioned in markets where demand is expected rather than simply increased across the network. For instance, TVS ILP expects a 10–15% rise in client volumes this year, particularly in consumer durables, electronics, lifestyle and automotive. Its festive preparation includes flexible and temporary warehousing, mezzanine storage, improved dock utilisation, dedicated fleet-management areas and measures to streamline vehicle movement. It is also strengthening last-mile connectivity and traffic flow, while deploying temporary canteens, workforce amenities and enhanced safety measures during peak periods.

But at what cost?
E-commerce is creating a tension for logistics: customers want faster delivery, easy returns, real-time visibility and flexible fulfilment, while businesses must control transport, warehousing, inventory and reverse-logistics costs. That tension is reflected in market growth. IMARC estimates India's e-commerce logistics market was worth $19.54 billion in 2025 and projects it to reach $103.83 billion by 2034, representing a compound annual growth rate of 20.39%. Transportation accounted for 41% of the market in 2025, B2C logistics 56%, and fashion and apparel 26%.

Social commerce adds another layer. Mordor Intelligence estimates India's social commerce market at $29.27 billion in 2025 and projects it to reach $143.86 billion by 2030, with growing participation from shoppers in Tier-II and Tier-III markets supporting expansion. Clearly, more transactions mean more parcels, but also more opportunities for failed delivery, returns and inventory fragmentation.

The intelligence layer
Artificial intelligence is moving beyond customelir-facing applications to become an operational layer across fulfilment. Adobe's India consumer research found that 90% of Indian consumers surveyed had seen their online shopping experiences boosted by generative artificial intelligence, particularly through convenience, personalisation and customer support. Google and Deloitte's research also identifies AI as a force expected to reshape commerce by 2030 through predictive logistics and automation.

For logistics operators, the real test is operational: can AI identify a bad address before dispatch, predict a failed delivery, select the right courier, position inventory closer to demand or determine when a returned product can re-enter stock?

But, e-commerce fulfilment no longer ends at the doorstep. It begins with forecasting and inventory positioning and runs through picking, packing, transportation, delivery and exception management. When an order comes back, inspection, grading, refurbishment and inventory recovery become part of the same chain. Across Emiza, Blue Dart, TVS ILP and Shiprocket, the focus is increasingly on eliminating unnecessary movement, reducing inventory idle time and improving successful delivery.

As India's e-commerce market heads towards the projected $250 billion by 2030, the competitive edge will lie not simply in moving parcels faster, but in moving inventory smarter: from demand to delivery, and, when necessary, from return back to saleable stock.