New Delhi, Dubai, New York City
The Brief
India is in its Cambrian moment for brands. The barriers that throttled brand creation in the West for fifty years (shelf, distribution, mass media) have collapsed. McKinsey projects $2 trillion in new consumption revenue by 2030, most of it homegrown. No existing discovery platform is built for the volume coming. The client came to us to build one, and to architect it to fork cleanly into new geographies from day one.
The Positioning
Most directories are utilitarian. Search bars, filters, grids. We rejected that. In a market flooded with brands, the discovery layer itself has to feel like a brand. If the container is generic, the contents disappear. If the container is confident, the contents feel curated. IBD was positioned as a sensory experience first, a data-base second. Closer to a museum than a marketplace.
The discovery layer is the brand. Everything else is content.
The Mark
The logo is a barcode. I, B, and D rendered as vertical lines. A barcode is the atomic unit of a brand: the line where an object becomes a product and a product becomes a purchase. By making the mark itself a barcode, IBD signals what it is without explaining it. The system scales from icon to billboard to a ceramic pod on a grey background. Even a rock can become a brand.
Three letters today. Any three letters tomorrow.
The System
IBD is not a website. It is a pipeline. AI agents scan the internet for new Indian product launches, ingest the information, and flag candidates for inclusion. Quality and relevance are filtered automatically, at the volume the market is producing.
.
Selected products move through an automated visual pipeline. Each image is stripped to a neutral background, re-composed in the IBD visual language, and converted into a subtle motion asset built for a scrolling feed. The catalogue looks curated even when it is automated. The machine does the volume. The aesthetic does the curation.
Three letters today. Any three letters tomorrow.
The Product
The site has three views. The first is an infinite plane. Users move in any direction and products surface in real time. No filters, no search, no visible taxonomy. Each product links out with a tracked parameter so the directory measures click-through and begins to read taste signals.
.
The second is a scroll view. One product fills the screen and changes as the user scrolls, one brand at a time. Slower, more editorial, built for users who want to sit with each product. The third is the submission layer, where brands and users add products directly. Every submission is tagged across vertical, category, price band, and availability. A structured dataset, built without the user ever seeing the structure.
Three letters today. Any three letters tomorrow.
The Tech
IBD was built in two phases. V1 was for speed: get a proof of concept live, generate real usage, learn from actual behaviour before committing to custom infrastructure. The front end runs on Framer. The AI layer combines Claude for reasoning, Gemini for visual processing, and custom logic for the handoff between discovery, processing, and publishing. V2 is being built now, every decision informed by real V1 data. Nothing in V2 is being built on assumption.
Every layer is replaceable. Every layer is API-driven.
The Fork
Geographic forkability was a requirement from day one, not a retrofit. The barcode mark accommodates any three-letter market code. The visual language holds across geographies with minor tonal adjustments. The tech stack is copy-paste deployable. A new country directory can be stood up in weeks once the local discovery agents are tuned to the market. Each expansion is tested before it is committed, and the learnings from each new market feed back into the parent system..
Three letters today. Any three letters tomorrow.
The Long Game
The current product is the foundation. The next layer is personalisation. Users log in, save products, follow verticals, and build taste profiles. The platform learns what a user responds to and surfaces brands they would never have found on their own. A parent in the first trimester gets kidswear months before they search for it. A new apartment owner gets homeware the week they move in.
.
The layer after that is commerce, reimagined. Agentic commerce sits at the centre: a logged-in user, a saved address, real-time inventory, and the discovery-to-purchase loop collapses into a single tap. The system handles pricing, shipping, and stock agentically across retail partners. Beyond that sit other models: limited drops, curated scarcity, event-driven discovery. Designed for the attention patterns of a market that has moved past the infinite scroll.
Discovery today. Commerce tomorrow. Taste, compounding.
The Economics
Most directories monetise through featured placement. Pay to be seen, pay to be ranked. We steered the client away from that model entirely. Paid placement corrodes trust in a discovery platform. The moment users suspect the surface is influenced by the wallet rather than the algorithm, the brand loses its most valuable asset: credibility as a taste-maker. The working thesis is a platform economy in three layers: a discovery layer that stays free and editorial, a commerce layer that captures value at the transaction, and a data layer that eventually becomes its own product for brands looking to understand taste at scale. The platform pays for itself without compromising what users come to it for. This is a Zero to One engagement in progress. We will keep building.
Trust is the asset. Everything else is downstream.
“The team at Lazy Eight think a few steps ahead of everyone else, on design and on tech. They have been co-founders on this from day one, not a studio we briefed and waited on. We will be building IBD with them for the next several years, and I cannot imagine doing it any other way.”
-Co-Founder, IBD