How Fynd helped Reliance Retail unify 380M+ customer identities across dozens of businesses

One identity, every Reliance Retail business: Marketing OS

The challenge

Reliance Retail runs dozens of separate storefronts, apps and physical formats - JioMart, AJIO, Tira, Reliance Digital, and many more - each historically capturing its own version of a customer. A single shopper could show up as an unconnected identity on five different platforms, with no unified view of who they are, what they've bought, or how they engage across online and offline. Marketing teams, meanwhile, were assembling their own monthly picture of performance by hand: stitching together weekly-business-review decks, finance trackers, and separate MarTech dashboards just to answer a basic question like whether a vertical was on plan.

Brand teams had a parallel problem. Understanding how a brand is actually perceived against real competitors meant pulling together funnel data, imagery surveys, usage numbers and ad-recall results from entirely separate sources, with no single place to see it all side by side.

The solution

A single identity layer with a Marketing OS being built on top of it as an agentic layer that answers the questions vertical and brand leaders actually ask. The foundation moves fast - consent-compliant identity, audience data and campaign activation, all streaming from real-time store and online data rather than static reports.

One identity across the whole portfolio

More than 380 million identities are already unified across Reliance Retail's online and offline formats, with over 200 profile attributes tracked per customer.

Store activity lands in minutes, not days

1.5 million daily orders from in-store POS stream into the platform from thousands of stores across dozens of formats, landing within minutes of checkout.

Built for scale and compliance together

The platform runs on Jio Cloud, is fully privacy-compliant, and holds sub-second performance even at that volume.

How it works

For a vertical or brand leader, a typical session runs through the platform's cockpits:

1. Open the Marketing Intelligence cockpit and see business outcomes, marketing spend and platform health for their vertical on one scrollable screen, benchmarked against plan rather than just against last year.

2. Drill into a Business Deep Dive if something looks off - a revenue story, a marketing-effectiveness breakdown, and an AI-generated diagnostic flagging what's actually driving the gap.

3. Ask a plain-language question directly - "why did average basket value drop in February?" - and get a charted, sourced answer without writing a query.

4. Switch to the Brand Health cockpit to see how the brand stacks up against a named competitor across awareness, consideration and usage, with the comparison brand swappable on the fly.

5. Check which brand-imagery attributes have moved most, and see what a piece of creative actually did for recall.

6. As the platform matures toward AI-Native UCP, the same leader will build a cohort, generate creative and launch a campaign from a single chat interface instead of switching between tools.

Under the hood

UCP runs as layers, from raw identity to an agentic marketing surface:

Identity

A DPDP-compliant identity layer with consent capture, a unified address book, one-tap retail single sign-on, and sanctions screening against a global watchlist - built API-first so any platform can plug in regardless of its own design system.

Data & Intelligence

A single, 360-degree customer view combining web engagement, offline store visits and retail KPIs, fed by an audience builder covering well over a hundred retail formats and real-time order streaming.

Activation

A campaign manager that runs audiences out to SMS, RCS, Meta and Google with governance built in - cooldowns, volume caps - tracks conversion through to ROAS, and closes the loop back into the data layer.

Agentic layer (Marketing OS)

Three cockpits sit here - Marketing Intelligence, Brand Health and the emerging AI-Native UCP - each aimed at a different question a leader asks, all pulling from the same underlying identity and data layers.

Still seeing your customer as five different people across five platforms?

Book a demo