Combining the segmented models



Hello, please help with the following scenario.

How can we bring the segmented models to same scale?

for e.g: A propensity model is to be built to get the likelihood of responding to a campaign.
I have come up with 3 segments based on customer profile and response rate and observed that response rate is different in these 3 segments. This lets us to build a separate model for each segment. Here the problem is probability of 0.3 in Segment 1 is not same as 0.3 in Segment 2 or Segment 3. How can I bring the probabilities of these segments to a single scale which can be comparable?