Key Findings
The marketplace is the store. 80% put Amazon and 71% put Myntra in their top three places to buy fashion online; 39% put a brand's own website there. Asked which platform they use most, only 14% named any brand's own site or app.
Shoppers like D2C brands but don't shop by the label. 25% say D2C brands appeal to them most, yet only 2 of those 19 named a well-known digital-first label as a favourite, and 17 of the 19 put Amazon in their top three. The brands shoppers name most are Zara (29%), H&M (28%) and Nike (22%).
Price opens the door; reviews and returns keep it open. 84% shop online for price and discounts. The top red flags are no customer reviews (76%) and no clear return policy (64%). Only 16% give influencer recommendations as a reason to buy.
The gap between photo and parcel is the core failure. 76% worry the product won't match its images and 75% worry about fit. Half of the bad experiences shoppers described were a different look, fit or item than the listing showed, and one in five involved a return or refund that went wrong.
Beating a low expectation earns a fan. 59% of the best experiences were about a product that turned out better than expected, and a quarter were low-priced items that beat their price.
Who Is Buying, and How Often
Everyone in this sample bought fashion online in the past month, and most buy often: 63% buy at least once a month and 23% buy every week. They also buy across the wardrobe. 86% bought two or more categories last month, and the average was 3.3.
Frequent buyers buy across the wardrobe. Weekly buyers bought 4.1 categories last month, against 2.4 for those who buy every two months or less. 76% of weekly buyers bought Indian wear (21% of the less frequent) and 65% bought formal wear (18%).
Sales and occasions set the calendar. 89% shop during sales, at celebrations, or both. Only 29% say they buy on impulse.
The Marketplace Is the Store
We asked shoppers to rank where they prefer to buy fashion online. Amazon leads clearly, with Myntra second. A brand's own website is ranked first more often than Flipkart, but makes fewer top threes.
A brand's own website divides shoppers. It is ranked first by 20% and last by 16% (12/76), and almost never second (3%, 2/76).
The ranking may overstate the brand website. When asked why they ranked it first, 5 of the 15 described a marketplace instead. Treat 20% as a ceiling.
The unprompted answers point the same way. Asked which platform they use most often, shoppers named 2.3 on average, and one in three named just one.
Amazon wins on trust; Myntra and Flipkart win on price. 45% of shoppers who rank Amazon first (15/33) gave trust, authenticity or reliability as the reason, against 22% of those who rank Myntra first (4/18). Price was the reason for 72% of Myntra-first shoppers (13/18) and 5 of the 7 Flipkart-first shoppers.
“Amazon has the best customer service in case there is any issue with the product, so that makes it more reliable for buying branded and expensive products.”
“It has a lot of choices, multiple brands in one place. I can compare prices and look for the best one.”
“It is a trusted website. I have been ordering from it for around six years.”
This matches our September study of avid online shoppers, where the marketplace often took the credit for a D2C purchase.
"D2C" Is Not How Shoppers Think
The D2C label does not predict D2C behaviour. Of the 19 shoppers who chose D2C brands, 5 ranked a brand's own website first and 4 ranked it last. 17 put Amazon in their top three, and 17 named only marketplaces as the platforms they use most. Only 2 named any of five well-known digital-first fashion labels (Snitch, Bonkers Corner, The Souled Store, Bewakoof, Nobero) as a favourite.
It works the other way too. Of the 16 shoppers who did name one of those five labels as a favourite, only 2 said D2C brands appeal to them most. Half (8/16) said they have no preference.
Global labels own the memory. 59% (45/76) named at least one international label such as Zara, H&M or Nike. 21% (16/76) named one of the five digital-first labels.
Marketplace house labels compete for the same memory. Myntra's house label Roadster was named as often as Snitch, The Souled Store or Bonkers Corner. 13% (10/76) named a Myntra house label or simply "Myntra brands", and 8% (6/76) named a platform such as Myntra or Meesho as a favourite brand.
“I prefer quality over brand, getting all the information about the product and the material used. I don't trust any brand blindly because it's famous.”
Brand names were typed freely; we grouped spelling variants. A few shoppers listed beauty brands. Grouping brands as international, digital-first or house labels is our own classification.
Price Opens the Door
Price gets shoppers online; returns are the next reason. Two in three name easy returns and exchanges, as many as name convenience. Shoppers chose four reasons on average, so price is rarely the only one.
Influencers rarely close the sale. Only 16% give influencer or celebrity recommendations as a reason to buy fashion online, the least-chosen reason by a wide margin.
“It was super cheap. I thought the quality would be good because a creator recommended it, but the shirt was torn and the buttons were coming out.”
The Photo Is the Promise
We then asked each shopper to describe one bad experience ordering fashion online, and coded the answers. The stories confirm the two concerns at the top of the list.
Half the bad stories are a gap between listing and parcel. 50% (38/76) described a product that looked different, fitted differently or was not the item ordered. Colour alone came up in 13% (10/76): a shade that differed from the photo, the wrong colour sent, or colour that faded.
“Getting an item that looks vibrant online but dull in person, or the fabric feels cheap and scratchy compared to the product description.”
Sizes don't hold, even within one store. Shoppers describe sizes that change from product to product and from brand to brand.
“Sometimes M or L sizes vary from product to product on the same site. One product in size M may not be the same size as another product in size M.”
“I once ordered a trouser with waist 42 but received a waist 46 labelled waist 42. The brand had a different size chart from other brands.”
Returns are where trust breaks. 21% of the stories end with a return or refund that was refused, failed or dragged on. A good exchange process can erase the problem: one shopper had never had a bad experience for exactly that reason.
“I returned it and they said they would refund the money into my bank account. I reached out to customer service multiple times but never got the refund.”
“I have never really had a bad experience ordering online, because if there is any issue with the size or quality of a product it is extremely easy to exchange it.”
What Stops the Order
No reviews, no sale. Three in four would not complete an order without customer reviews.
Payment trust stops half of buyers. 52% stop at an unfamiliar payment gateway and 44% at no cash on delivery. COD is not only a low-income ask: 56% of shoppers reporting incomes of ₹3 lakh or more flag it (18/32), against 35% of those under ₹3 lakh (15/43).
“Most of the time the platform does not have cash on delivery, and the order got cancelled without my consent.”
The Surprise That Earns a Fan
We also asked each shopper to describe a time a fashion order turned out far better than expected.
The good stories mirror the bad ones. 59% (45/76) of the best experiences were about quality, look, fit or durability beating expectations. The best stories are about the product, not the delivery: only 8% mention speed.
Low expectations are the opening. A quarter of shoppers described a cheap item they expected to disappoint that turned out well. Three shoppers said a surprise like this led them to recommend the product or the seller to friends.
“It turned out way better than my expectations, as it was the same as shown in the image, which is quite rare nowadays while shopping online.”
“The price was around ₹400 and I thought it would be good for layering only. Surprisingly, it was very warm. It was a good purchase and I still use it to this day.”
“I was blown away when I first received them, and recommended it to all my friends.”
New brands can win this way, but rarely do yet. Only 3 of 76 described a lesser-known or untried brand that delivered.
“When I ordered from an affordable, lesser-known online brand, the fit and quality were better than expected.”
“I ordered a shirt from a brand I had never heard of. I bought it because of the offer and took my chance, and it turned out to be good.”
Who Differs
Women lean harder on Amazon and Meesho, and worry more about returns. Almost every woman in the sample puts Amazon in her top three. Fewer put a brand's own website there.
The most frequent buyers want unique collections but find them on marketplaces. 88% of weekly buyers shop online for unique collections, yet only 3 of 17 put a brand's own website in their top three.
Non-students are stricter about service. Non-students are more likely to stop at an unclear return policy and to worry about customer service.
Shoppers who rank a brand's site first want something different. 80% of the 15 (12/15) buy online for unique collections and 80% for quality and reviews, against 61% and 66% overall. Fewer name price (67%, 10/15, against 84%). They care less about missing reviews (53%, 8/15, against 76%) and more about a clear return policy (80%, 12/15, against 64%).
Subgroups are small, and 5 of the 15 brand-site-first shoppers described a marketplace when asked why. Read every difference in this section as a signal, not a measurement.
The Playbook for D2C Fashion Brands
Sell where shoppers already shop: 80% put Amazon and 71% put Myntra in their top three. Only 14% name any brand's own site or app among the platforms they use most.
Make the photo match the parcel: 76% worry the product will not match its images, and half of the bad experiences were a different look, fit or item than the listing showed.
Keep one size chart across the range: wrong size or fit is the most common bad experience (28%), and shoppers report sizes that change from product to product.
Make returns work end to end: 64% will not order without a clear return policy, and 21% of bad experiences were returns or refunds that were refused, failed or dragged on.
Seed customer reviews first: no customer reviews is the top red flag (76%).
Offer COD and a payment gateway buyers know: 44% stop at no cash on delivery and 52% at an unfamiliar gateway.
Beat the price expectation: a quarter of the best experiences were low-priced items that turned out better than their price.
Leading with "D2C" as the pitch: only 2 of the 19 shoppers who prefer D2C brands named a well-known digital-first label as a favourite.
Counting on your own website for the first order: 17 of those 19 named only marketplaces as the platforms they use most.
Influencer endorsement as the reason to buy: only 16% give it as a reason to buy fashion online.
Competing on price alone: shoppers who pick Myntra or Flipkart first already pick them for price, and shoppers who rank a brand's site first cite collections and quality (80% each) more than price (67%).
Methodology
84 people started an AI-led survey on SpillBill between 24 October and 1 November 2026. Four were screened out because they had not bought fashion online in the past month, three did not finish, and one completed response was excluded for a low quality score. This paper analyses the remaining 76, all of which passed quality checks.
The sample spans 31 districts in 15 states, with 24% in Bengaluru Urban. It is 71% male, 87% aged 18–24 (median age 20) and 70% students, and 57% report incomes under ₹3 lakh. It skews young and male, so read subgroup findings as directional.
Three responses had one answer recorded against the wrong question. We excluded those answers, so the frequency, brand-type and red-flag questions have 75 responses. The ranking question asked for a top three; we report the first three positions. Open-ended answers (platform used most, reason for the top platform, favourite brands, one bad and one better-than-expected experience) were coded by hand, and one answer can carry several codes.
This paper names platforms and brands only as shoppers named them, and names no respondents. Quotes are lightly edited for grammar, with platform and brand names removed from quotes about bad experiences. Every finding describes a pattern in this sample, not a proven cause.