Using models makes one a better thinker.
peer model and
percolation model some thresholds in spreading information through graph networks (including social) are needed for starting a chain reaction. A lot of projects may fail completely only because they lacked a wee bit to overcome the threshold and because they are bad designed. Sometimes a little may be needed to transform fail into a success.
Sometimes models drift between branches of science and they may have their births in one science, be improved in another and than come back "home" but in a new quality, enriching their parents.
There are a bunch of effects and models, some of them may work together, some not. Usually they are
Several models that may (or may not) work: finding the bottle neck, aggregation model, peer effect model (chain reaction), leverage (butterfly), etc.