It's difficult, but not impossible, and here's how... I could give you at least five case studies from the past year in which specific things I did created thos

It's difficult, but not impossible, and here's how... I could give you at least five case studies from the past year in which specific things I did created those kinds of gains / savings for a client. It's basically standard CMA stuff - show comparable properties, where they sold, where we sold (or had our offer accepted) and demonstrate the tactics that got to that point. Do that across a group of agents and you quickly have a comparative model. Now - sometimes there are emotion-driven data outliers. For example, sometimes you will see someone overpay in cash for a property that could not have appraised at that same value. So again, using data you can account for those scenarios as well. Point being - all that sales data is available and you can pretty accurately create a model of which agents overall are having success not only selling properties, but the degree to which their clients are winning negotiations. I'd argue that even if an agent was being "overruled" by their clients - even that would be interesting, as it would show that their clients were not actually listening to the negotiating advise offered by that agent (IE: the data could actually demonstrate that there are communication / trust issues between the agent and their clients).