[NirDremer]

That's a great question.

The relatively easy part is to find who recently raised money and we do it mostly through monitoring the press and limiting what we present only to companies that fit the criteria.

Behind the scenes we do much more complex stuff in order to evaluate the companies and a hint for that can be seen in the badges below the company description. In essence, we built search engine that profiles companies in depth by looking at the companies website, github activity any anywhere we could find information about the founders, investors, team, product and even the code itself.

This is being developed mostly for our discovery engine - Yodas (www.yodas.com) and we benefit from it for Hyper.

[jstandard]

Do you have a way to backtest the algorithm? Would be interesting to be able to verify accuracy without needing to wait 1-2 years (or whatever the time period meant by "about to go into hyper growth").

The reason I ask is I tend to be pretty cautious about these types of predictors of "business success" given that VCs nor research teams ("Good to Great" by Jim Collins) don't exactly have a strong track record.

An interesting test would be to make it time bound and run it on something like Clinkle circa late 2013.