The Viral Economy In A Nutshell

Estimate the potential of a video concept (we call it “Pricing”) through the title and thumbnail by giving it a value in views at about 3 months. It’s not a direct view prediction, but they are closely related (see how here).
Simulate an A/B test to rank different versions of thumbnails and titles. Test infinite combinations for the same project.
Evaluate why a video performed poorly: was it the concept or did the production fail to match viewer expectations?

People vote daily on others’ projects. Each project requires hundreds of votes to complete and our algorithm to clean the noise on the top of it. For a “Pricing”, the entire process takes about 48 hours to complete. A/B simulations may take 24 to 72 hours, depending on how many combinations you're testing.
The voting system is managed by an internal economy to keep vote quality high (learn more here).

It doesn’t measure watch time. People who vote have no idea if a piece of content is remarkably well produced or not, which can skew the results significantly depending on the gap between expectations and the actual content.
It doesn’t work well for niche or non-English content. If specific knowledge or context is needed to understand your video, the tool might struggle (learn more here).
It doesn’t scan YouTube. Our algorithm is connected to nothing else but the data we feed it with, it is completely blind to everything else.