There is an enormous amount of investment research into both "earnings surprise" and "estimate revision". Entire companies are built on the concepts, so far more than can be explained in this little box.
Short (and overly simplistic version): Companies with positive surprises tend to have more positive surprises. Negative surprises tend to lead to more negative surprises.
I have not personally tested Zacks. It is generally a reputable company, but those claims seem quite high.
Revision and Surprise are best applied across a diversified portfolio. It is possible to trade individual stocks on the info, and used to be fairly easy money, but it has become much more difficult as estimate data is now so widely available. New information is reflected in stock prices quicker and more fully than in the past.
