Where founders and community owners launch prediction markets. @shiftmarkets
295 Madison Avenue NYC 10017
Joined June 2026
- Tweets522
- Following660
- Followers319
- Likes982
Pinned Tweet
Prediction markets hit $1.2 billion a month in early 2025.
By May 2026, that number reached $28 billion.
Robinhood alone traded 8.8 billion event contracts in Q1. That’s the revenue power prediction markets has.
Nothing in finance is moving faster, and most of the markets that will define this category don’t exist yet. 👇
NEXTPredict kicks off tomorrow at Hudson Yards, the first summit built entirely around prediction markets.
That alone says a lot about where this space is heading.
The Shift team will be there. See you at Hudson Yards!🗽
A bug bounty of up to $5M, six security reviews, and nearly four weeks of live testing.
That’s what a rebuild of the contracts under a market looks like.
Polymarket is set to switch new markets to V2 on Nov 2, and each one can resolve through UMA, Chainlink, or a manual oracle. 🔧
The playbook here is clear:
Build the audience with one product, then launch the next one on top of it.
Polymarket Perps just passed $4.5B in 30 days, and Kalshi just added a US 500 perp.
No human reviewers, 179,842 past Polymarket markets, and GenLayer says every market it resolved matched the real outcome.
Resolution is the costly part of launching a market, so cheaper resolution means more markets on small questions, and more room for builders.
Ambiguous ones are the real test.
We ran GenLayer on 179,842 past Polymarket markets. No price feeds, no human reviewers, and every market it resolved matched the real outcome.
@EdgarsNemse with @letjaybuild on why prediction markets are just the start.
Full interview below 👇
A study of internal prediction markets at Google and Ford found the same flaw in both: employees were too optimistic.
Your team’s launch date probably has it too.
Shift builds the infrastructure to launch a prediction market in weeks. 🔧
Launching a prediction market used to mean building a trading venue from scratch.
Now matching, settlement and compliance can come pre-built.
The hard question left is which community cares enough about the outcomes to use it.
Steal this startup 💡
Hyperlocal prediction markets.
Will the new bridge open on time? Will the corner shop survive the year? Will they finally fill that pothole?
Nobody prices the stuff people in your town actually argue about.
The infrastructure to launch markets like that already exists. What's missing is someone who knows the community and how to reach them.
A few weeks ago we said someone should build the layer that sits across every prediction market.
@ParlayXOfficial is doing it for institutions. One account routes orders across Polymarket and Kalshi, with more venues on the way. Its founder calls it Plaid for prediction markets.
The aggregation layer is getting built. It’s early, and there’s room for more than one.
Steal this startup💡
Right now, serious prediction market traders keep five apps open and tab between them by hand to compare odds. No one has built the tool that does it for them.
Kalshi and Polymarket keep posting JUST IN odds as breaking news, which doubles as free marketing for them. But it also means the live data is all out there in the open, waiting for someone to pull it into one screen.
The terminal that reads every platform at once is an open lane. Whoever builds it owns the layer all of them depend on.
Hot take: nobody has ever picked a prediction market because of its matching engine. 🔧
Rent that part. Spend your time on building your audience.
TOKEN2049 Singapore just wrapped.
Organizers say 25,000 people came through, and the biggest change from last year was who showed up:
Goldman, BlackRock, Nasdaq and Morgan Stanley all had people in the room.
What it means if you are building🧵👇
Prediction markets had a seat at the main table too.
Polymarket's CEO was on the main stage, Kalshi talked about bringing perps to the US, and one panel was literally titled "When Everything Becomes Tradable."
The category is in the room now. The opening is building what sits underneath it.
$84.6B traded across prediction markets in the last 30 days, per TickerTracker.
Kalshi says it doesn't make money when a trader loses.
Launch your own market and the same thing applies to you.
Your job is getting people to trade. Who wins isn't your problem. 📊
Polymarket $POLY chatter is back 🧵
If a token and airdrop land, the interesting question is what happens after everyone checks their allocation.
Here’s our take on what it could mean for prediction markets 👇
The catch: reward volume alone and you could get people trading for the airdrop instead of expressing a view.
The activity chart looks great. The forecasts might not improve.
Prediction markets need incentives that reward useful participation.