The Cross-Store Identity Network: Autofill Everywhere

I want you to sit with that Nagpur example for a second, because it is the whole idea in one scene. A first-time buyer on your store, and yet her address fills itself. She types ten digits, confirms a code, and the house number, area, pincode, city — all of it appears. To her it feels like magic. To the store it feels like a conversion that would otherwise have leaked at the address step. It is neither magic nor luck. It is a network.

The single most underrated fact about Indian ecommerce is that the mobile number is the real identity. Not the email, not a cookie, not a login. The phone. The same number follows a shopper across every brand she buys from. And once you accept that, a powerful idea falls out of it: if many stores all recognise buyers by phone, and they quietly pool what they know about each verified number, then any one store can greet a returning buyer as if she were a regular — even on her very first visit.

What the network actually is

Strip away the jargon and it is simple. Every time a shopper completes an order on any store in the network, that order produces a durable fact: this phone number ships to this address. Multiply that across hundreds of brands and millions of orders, and you have a living map of where India's online shoppers actually live and how they write their own addresses.

No single store could ever build this alone. A mid-size Shopify brand sees a buyer maybe two or three times a year. That is not enough to know her. But she shops across dozens of brands, and collectively those brands see her every few weeks. The network turns that scattered, per-store trickle into one continuous picture. Your store contributes what it learns and, in return, draws on everything every other store has learned. That trade is the entire deal, and it is why this compounds in a way a standalone address autofill feature never can.

This is exactly the moat that GoKwik and Shopflo have been quietly building. It looks like a checkout feature from the outside. Underneath, the real asset is the identity graph — the accumulated knowledge of who buys where and ships to which address. A rival can copy the checkout UI in a month. They cannot copy years of pooled order history overnight. That is what makes it defensible.

Why more stores make it better for everyone

Here is the part that makes it a network and not just a database. The value each store gets out is a function of how much everyone else has put in. When a new brand joins and starts contributing orders, it does not just take — it feeds fresh addresses and fresh confirmations back into the pool. Every store's autofill gets a little sharper. Coverage widens into more pincodes, more tier-2 and tier-3 towns, more of the buyers a single brand would never see enough of on its own.

Economists call this a positive network externality. In plain terms: the tenth store benefits the first store, and the thousandth store benefits all nine hundred and ninety-nine before it. Autofill hit-rates climb for the whole network at once. This is the opposite of a normal software feature, which delivers the same fixed value on day one and day one thousand. An identity network is worth more every month simply because more orders flowed through it, and none of the existing members had to lift a finger for that improvement.

The privacy model is the whole design

The obvious worry, and the right one to have, is privacy. Are we passing people's home addresses around between random stores? No. And the DPDP era in India means you cannot hand-wave this. The design has to be privacy-safe at its core or it should not exist.

Two things make it safe. First, the phone number is stored as a hash, not in the clear. The network matches on a fingerprint of the number, so no store is holding a browsable list of everyone's mobile numbers and where they live. Second, and this is the important one, the address never reveals itself just because someone typed a phone number. It reveals only after the shopper passes an OTP that goes to her own phone. She is the key. If she does not enter the code that landed on her device, nothing unlocks.

That single rule closes the abuse case. A curious stranger cannot type your number into some store and read your address, because they do not have your phone in their hand to receive the OTP. Consent is not a checkbox buried in a policy — it is the physical fact of holding the device the code was sent to. The buyer proves she is herself, and only then does her own data come back to help her. It is her information working for her, exactly when she asked it to by choosing to check out.

The sibling network: reputation for returns

Once you have an identity graph keyed on phone, a second, quieter network becomes possible, and for COD-heavy India it might matter even more than autofill. If the network knows the addresses a number ships to, it can also know how those shipments ended. Delivered. Returned. Refused at the door. Repeatedly undelivered.

That turns into a shared reputation. A buyer who refuses cash-on-delivery parcels across five different brands is a known risk the sixth brand can see the moment she checks out — not because that store has any history with her, but because the network does. This is the backbone of serial returner detection, and it flips the economics of RTO. Today every store eats the same expensive lesson alone: ship, get the return, lose the shipping both ways, learn nothing that carries forward. In a network, the lesson is learned once and shared. The reliable buyer sails through with COD open, and the serial returner sees an OTP-gated COD or a prepaid-only nudge. Pair that with a real RTO reduction playbook and the return rate on the whole network trends down over time.

Same privacy rules apply. No store sees another store's raw order history. What it sees is a risk signal derived from the network — a score, not a dossier — and only for a number the buyer herself just verified. The buyer's good behaviour becomes an asset she carries from brand to brand, which is genuinely fairer than the status quo where every honest COD buyer is treated with the same suspicion as a chronic returner because no store can tell them apart.

Why this compounds, and what it means for you

Most tools you bolt onto a store give you a fixed lift. A faster page, a better button, a coupon — useful, but the gain is roughly the same the day you install it as a year later. An identity network is different in kind. It gets better on its own. Every order that flows through it, on any store, anywhere, makes the next buyer's checkout a little smoother and the next risk call a little sharper. You did nothing and it improved.

Play that forward. In two years, a network that started thin is covering a large share of India's D2C shoppers. Autofill stops being a nice surprise and becomes the expected baseline — buyers will feel the friction of any store that makes them type an address from scratch, the way we now feel friction typing a card number when UPI exists. The brands inside the network ride that rising tide. The ones outside it look slower and more suspicious every quarter, through no change of their own.

So the practical takeaway is not really about a feature. It is about which side of the compounding curve you want to be on. A cross-store identity network is a one-time decision to join something that keeps paying you back — faster checkouts today because the graph already knows your buyers, and a widening RTO defence that improves while you sleep. The one-page checkout is the visible surface. The network underneath is the thing that quietly gets harder to compete with every single day it runs.