A contrarian approach to building the next great prediction market platform
Everyone's trying to build the "Bloomberg Terminal for prediction markets" – sprawling dashboards that aggregate thousands of markets across dozens of topics. But what if we're thinking about this all wrong?
My instinct about how this could potentially be cracked has to do with focus. Instead of starting out trying to aggregate many predictions about many events, what if we could get one really good page going for one single topic, or even one single important market, and then expand from there?
Picture this: the best page on the internet for understanding "When will the war in Ukraine end?" – assuming we have a market for that. Every time the market moves, AI does the heavy lifting of researching why it might have moved, building a digestible narrative and timeline from that movement.
This isn't just market data; it's market intelligence. When the odds shift from 15% to 23% for "war ends by December 2024," readers don't just see numbers – they get context. What news broke? Which analysts are talking? What historical parallels might apply?
(This would probably work brilliantly as a newsletter format too – imagine getting a weekly digest that not only tells you how the market moved, but why it moved and what that means.)
I've been experimenting with this approach myself. As a US Soccer nut, I built a single-page treatment for my prediction market about the USA World Cup roster. Instead of trying to cover every sport or every tournament, I went deep on one specific question that I (and hopefully other soccer fans) genuinely care about.
The result? A focused, rich experience that actually helps users understand not just what the market thinks, but why it thinks it. Users can see the narrative unfold in real-time as new information changes the odds.
The traditional aggregator approach faces a classic chicken-and-egg problem: you need lots of markets to be useful, but you need users to make markets liquid. By starting with one exceptional page, you can:
- Perfect the format without getting overwhelmed by scale
- Build genuine expertise in your chosen domain
- Create something people actually want to bookmark and return to
- Prove the concept before expanding horizontally
Think about how Nate Silver built FiveThirtyEight – not by covering every possible statistical topic, but by becoming the place for election forecasting, then expanding from that foundation of trust and expertise.
Once you've nailed one market, the playbook becomes clearer. You're not just adding more markets – you're adding more expertly curated markets. Each new topic gets the same deep-dive treatment that made your first page indispensable.
Maybe you start with Ukraine, then add the 2024 election, then climate targets, then Fed rate decisions. Each page becomes a destination in its own right, not just another row in a spreadsheet.
This approach acknowledges something important about how people actually consume information: we don't want to be overwhelmed with data, we want to be informed by intelligence. The difference between showing someone a market price and helping them understand what that price means is the difference between a tool and a service.
In a world where everyone's building horizontal platforms, there might be real value in going vertical first. Start with one market, make it exceptional, then expand from a position of strength rather than trying to boil the ocean from day one.
Sometimes the best way to aggregate everything is to start by perfecting something.
What do you think? Are there specific markets or topics where this deep-dive approach could work particularly well? I'd love to hear from other folks pursuing interesting ideas in this space.