Measurement and Results - Preventing IS intensity drift

What should the intensity of the IS signal be to get the best stability? I am testing cannabis samples of multiple matrix types. I see a great deal of change in the signal of the IS for plant based samples compared to the other samples and QCs. They change as much as 500%! I have controlled it somewhat by adding up to 6% isopropyl alcohol to all samples. I have an in-house IS mix of Rh, Tb, Tl and Y; the ones I have found to drift the least I adjust the concentration so the signals of all 4 are 1500 to 2000 kcps. Should they be higher? Thank you.