How we helped Polymarket think about the experience of a platform the world was watching

In our engagement with the Polymarket team, we worked across two phases — first exploring the foundational problems of the homepage, then developing a full concept that made the scale and liveness of the platform legible at once.

The platform worked. The experience hadn’t kept pace.

Polymarket is the world’s largest prediction market — a place where people put real money behind their beliefs about global events. Presidential elections. Fed rate decisions. Crypto prices. Premier League outcomes. By the summer of 2024, it had become one of the most cited sources of real-time political sentiment in the world. Millions of dollars in active positions across hundreds of markets. The product had validated its model. The problem was that the interface didn’t communicate any of it.

Prediction markets live and die by two things: urgency and community. The sense that something is happening right now, and that other people are already in it. A platform can be extraordinary in scope and still feel inert if the experience doesn’t surface that energy. Polymarket’s did.

Starting with the right questions.

Before any design work began, we needed to understand what the homepage was actually for. Not functionally — Polymarket's team knew their product well. But experientially: what should a user feel when they land on the page? What should they know immediately, without having to look for it?

Three things. That the platform is alive. That there is something worth paying attention to right now. And that other people — a lot of them — are already engaged.

Those three goals became the framework for everything that followed.

Phase one: exploring the problem space.

In August 2024, we delivered the first phase of work — concepts and wireframes focused specifically on homepage activity. This wasn’t polished design. It was structured thinking, made visible.

We explored three distinct directions.

The first was layout — how the page should be organized to serve both new and returning users. New users need orientation: a clear value proposition, a path into the platform, an understanding of how it works. Returning users need none of that. They need their active bets, their watchlist, and the fastest possible path to the markets they care about. The same page has to do both without compromising either. Our proposed solution was a state-aware layout — the page adapts based on whether a user is logged in, foregrounding onboarding for newcomers and active positions for those who are already in.

The second direction was dynamic content — the mechanisms that make a page feel alive rather than static. We explored rotating carousels that cycle through trending markets every few seconds, event countdown timers that create urgency around major upcoming events, live trading stats that display real-time volume and activity, and animated market cards with live trade notifications overlaid directly on the content. None of these were decorative. Each was a specific answer to the question: how does the interface communicate that something is happening right now?

The third direction was social proof — the signals that tell a user this platform has an active, engaged community behind it. A live ticker of user trades, profits, and ROI. Market sentiment indicators — likely, live, overold — that give a user an immediate read on crowd behavior. A leaderboard in the hero section surfacing top performers, their positions, and their returns in real time. These were the elements that would make Polymarket feel inhabited, not just functional.

The tension we were designing through.

The challenge with all of this was noise. Adding dynamism to a page is easy. Adding dynamism without overwhelming the user — without turning a discovery surface into a slot machine — is the harder problem.

Every dynamic element we proposed had to pass a single test: does this help the user make a decision, or does it just stimulate them? The countdown timer helps. It tells you something is about to resolve and you need to act. The live trade ticker helps. It shows you what informed users are doing with their money. The sentiment badges help. They give you the crowd’s position at a glance. Animated flying coins do not help. They stimulate. We removed them.

The same logic applied to the leaderboard. Surfacing top performers creates aspiration and competitive energy — but only if the information is real and specific. Placeholder data undermines it. We designed for the populated state, with actual usernames and actual returns, because that is what creates the feeling the feature is meant to produce.

Phase two: a full homepage concept.

In September 2024 we delivered the second phase — a complete homepage concept, informed by the August exploration and grounded in competitive research.

We studied how platforms operating at content scale and high engagement handle information density. Medium, Airbnb, Netflix, Facebook. Each had faced a version of the same problem — more content than any one user can consume, an interface that has to make intelligent choices about what to surface and when. Each had arrived, through iteration, at a similar answer: visual hierarchy does the work that filters cannot.

Larger cards for critical content boost engagement because users spend the majority of their time above the fold. Consistent visual systems improve retention. These were the arguments for the direction.

From that foundation we designed the full experience. The homepage was divided into three functional zones: navigation, discovery, and activity. Navigation gives users who know what they want a direct path to it. Discovery surfaces what’s new, trending, and relevant. Activity makes the platform feel inhabited — a live feed of what other users are betting, in real time.

The featured card system sat at the top of the page — large-format, custom-designed cards for major events, intended to change frequently and carry the visual weight of the moment. The 2024 US Presidential Election. Elections Forecast. Premier League Countdown. Each card was an editorial decision as much as a design decision. The featured section had to feel like a front page — because for millions of users during the election cycle, it was.

Below it, the prediction card system was rebuilt from the ground up. Single-outcome cards show probability and allow direct action from the card surface. Multi-outcome cards surface the top five, with the rest hidden. Hover states reveal bet actions without cluttering the resting state. The distance between seeing a market and acting on it was reduced at every point.

The activity feed addressed the liveness problem directly. A persistent real-time stream of user bets, volumes, and positions — visible on larger screens, collapsible on smaller ones. It had to surface the feeling of other people making decisions alongside you. Prediction markets are inherently social. The activity feed made that visible.

Search was redesigned as a discovery surface. Opening it surfaces three immediate categories — For You, Trending, New Markets — before a query is typed. Personalized navigation adapted to each user’s history. Onboarding cues were placed unobtrusively for new users, without interrupting the experience for those who were already in.

The engagement closed with a fully interactive prototype — every component built into a coherent, testable system, ready for the Polymarket team to validate before committing to build.