The Mobile Checkout Problem Nobody Talks About
Why Indian mobile checkout still leaks orders between cart and OTP, and what to actually fix instead of blaming the payment gateway.
Mobile checkout on Indian storefronts is not slow because of network speed. It is slow because we ask people to type a full address on a 6 inch screen while a UPI collect request quietly times out in another app. Most orders leak between add to cart and the second OTP screen, and most founders never watch a real user do it. If you have not sat next to someone completing a Rs. 899 order on a mid range Android on patchy 4G, you do not actually know your checkout.
The leak is not where you think it is
Founders love blaming the payment gateway. Razorpay this, Cashfree that, PayU took three seconds. In reality most abandoned mobile carts in India die before the gateway even loads. They die on the address form. Pincode field asks for six digits, then silently fails because the user tapped a space. State dropdown has 36 entries in alphabetical order and the thumb has to travel the whole screen. Phone field rejects the +91 the user pasted from WhatsApp. By the time the buyer reaches the pay button they have already decided to try again later, which means never.
Watch this on a Redmi Note with one bar of Jio in a Tier 2 town. It is not the experience you see on a MacBook simulator. The keyboard covers half the form. Autofill is inconsistent across Chrome and the Instagram in app browser. If the buyer came from a Reel, they are inside Instagram's webview, and Google autofill does not work there at all. Your checkout is running on a different browser than you think.
UPI collect is the silent killer
Everyone celebrates UPI share on their checkout dashboard. Nobody looks at UPI collect failure rate on mobile. The buyer taps Pay, gets redirected to GPay, misses the notification because they were reading something else, comes back to a timed out session. The order shows as failed. The customer thinks money was deducted, opens WhatsApp to complain, and now you are a support ticket, not a sale. UPI intent flow, where the app opens directly with the amount prefilled, converts far better than collect. Most Shopify stores in India are still on collect by default because nobody switched it.
COD is not a bug, it is a symptom
When a store's COD share is above 60 percent, the founder usually blames Indian buyer behaviour. It is not that. It is trust. The buyer does not trust that if the prepaid order fails, the refund will actually come back in three days. They have been burnt before, by a beauty brand, by a small D2C label, by somebody's cousin's Instagram store. COD is insurance against your checkout, not a preference. Fix the prepaid experience, add clear refund copy on the pay screen, show the payment logos above the fold, and COD share drops on its own. Then RTO drops. Then margins move.
What to actually do this quarter
Three things. One, record a real session. Not Hotjar aggregate, an actual screen recording of a customer on a cheap Android completing a Rs. 500 order. You will see the problem in 90 seconds. Two, switch UPI to intent flow, put it as the default option, and stop hiding it behind a Cards accordion. Three, cut the address form. Pincode first, autofill city and state, phone last. Remove the second address line unless the buyer taps to add it. This is not a redesign, it is a one afternoon fix that most stores never do because everyone is busy launching new SKUs instead of fixing the one page that decides revenue.
Common questions
Is a one page checkout always better than a multi step one?
Not always. One page works when the form is genuinely short, four or five fields, with smart pincode autofill and UPI intent as the default. It stops working when you cram address, shipping options, coupon, GST invoice toggle, and payment into one scrollable mess. On a mid range Android that is 1200 pixels of scroll before the pay button. Multi step is fine if each step fits above the fold on a 6 inch screen and has one clear action. Test the actual scroll length on a real phone before you pick a side.
Does shoppable video actually help with the checkout problem, or does it just add another step?
It helps only if the tag opens the product page in the same session and the buyer can add to cart without leaving. If the widget deep links out to a new tab and loses context, you have made things worse. beyondRegular carousels and reel feeds keep the buyer on your site, tag products in frame, and hand them off to your existing checkout with your existing gateway. The video does not process payment. It gets the buyer to the pay screen with intent already built, which is where most other traffic sources fail.
Should I turn off COD to force prepaid adoption?
No, not before you fix the prepaid experience. Turning off COD on a store with a leaky checkout just kills orders, it does not convert them. Fix UPI intent, cut the address form, add trust signals on the pay page, then run COD as a paid option with a small handling charge, say Rs. 30. Buyers who really want COD will still pay it, and everyone else will drift to prepaid on their own. That is the sequence. Reverse it and you lose the quarter.