JelloWorld puts a confidence score on content and on the people making it. You submit a URL, you get an assessment of how reliable it is. Creators build a score over time that travels with them.
To become the gold standard of online truth by providing a confidence score for reliable content and creators.
Only humans allowed
“While AI helps us process information quickly, only real humans make the final call on what’s trustworthy. Because truth matters too much to leave to algorithms alone.”
This is the product's whole position, and it's a harder one to hold than the alternative. Full automation scales and costs nothing. A human in the loop is slow and expensive — and it's the only version of a trust product that has any standing to call itself a standard.
Four things it
has to get right
Confident fact-checking
Verify information so you can confidently make important, life-altering decisions.
Build trust
Establish credibility with your peers and your audience. Foster a community based in truth.
Boost engagement
Share verified content and spark meaningful conversations with your audience.
Confidence scores
An assessment of content reliability from a particular creator, based on the verification tool.
Making verification something people want to do
Fact-checking is civic homework, and homework doesn't retain anyone. So the submission flow is built as combat — a structured, multi-step analysis with points, levels, and badges on the other side of it. The Truth Arena turns a claim into something you play against rather than a form you fill in.
The risk is obvious: gamifying truth can reward volume over rigour. The human-in-the-loop review is what keeps the scoring honest.
Become a JelloStarter
Score unlimited facts, convo starters, and link up with trusted content creators. Every piece and vetted creator comes with a confidence score so you always know what’s reliable.
One word covers a member of the platform and the conversation product — which is how the card game and the verification engine end up being the same brand rather than two.
Read the Conversation Starter Game case study→What it's built on
What it taught me
What I got right
Refusing to let the algorithm make the final call, and saying so on the homepage. It's the least scalable decision in the product and the only one that earns the phrase 'gold standard'.
The hard tension
A confidence score is a number, and numbers get trusted more than they deserve. The honest version of this product has to keep showing its work — where the evidence came from, who reviewed it — or it becomes exactly the kind of unexamined authority it exists to push back on.
What I would do differently
Ship the URL analyzer to a small group before building the marketplace, the points system, and the combat flow around it. The scoring engine is the product; everything else is packaging, and I built a lot of packaging early.