Relative Valuation Peer Tiers vs. Beta calculation

What if I have some peers–say 12–divided in three tiers depending on their degree of similarity: 4 tier I, 4 in tier II e 4 in tier III. I did that because I wanted to calculate mean multiple for my target company giving different weights to different tiers. (e.g. tier I 100%, tier II 75%, tier III 50%)

When calculating the beta should i use the same peer companies? If so, should i weight them according to their tier to be consistent?