In statistics, a mediator is a factor that sits between an independent variable and a dependent variable. They are the middlemen of statistical relationships. When a variable is a mediator, it is being influenced by an independent variable, and in turn influencing the dependent variable.
For example, in research, we often see relationships where there is a strong positive correlation between two things. In our B2B Buying Experience Research we saw that as buying group sizes got bigger, buying cycles got longer. But this relationship begged the question: Ok, but what causes the buying group size to increase? As marketers and sellers, we probably can’t do anything to influence the size of a prospect buying group, so knowing that is interesting, but not very useful. But, we suspected that there was something operating on buying group size. What causes them to be bigger or smaller? As we thought about that, we landed on the number of vendors being evaluated as a possible driver of buying group size. After all, if the buying group was going to have to do more work, perhaps they would increase the size of the team. So, we tested that, using mediation analysis.
Indeed, buying group size was a mediator that sat between Vendor Count (the number of vendors being evaluated) and Buying Cycle Length. As Vendor Count increases, buying group sizes also increase. The increase in buying group size causes buying cycles to increase.