@Aaron Carlock At a high level, we are essentially documenting every step of the sales process very carefully. This actually represents a lot of data over the years. As a basic example, for any given client or prospect, I could tell you exactly how many emails were sent before a decision was made, what the service offered is, and at what price point. We also have access to all property data associated with addresses. So one naive idea I have just starting out is to build a model that can take property data as an input and output a price point for a given service along with a confidence interval of the likelihood that the proposal would be accepted. Again, I am extremely new to data, so I am not quite sure yet what is possible, but at the very least I am excited to see what's possible with the data we have.