Evaluation of reliability of suppliers by on time delivery

I am trying to create a weight that can help me properly evaluate the reliability of a company’s on time delivery ratios. I’m using historic delivery data. For example, comparing company A that delivers 5 times a year, 3 times successfully. Is not fair to compare to company B that delivers 73 times a year, 72 times successfully.

I’ve considered analytics weights and have done similar projects on R studio with weighing sample populations. Lost on how to create a weight for successful on time deliveries in excel.

Using prior stats knowledge, I know you shouldn’t predict something with such a small sample size. Is there any topics I should research on how to create a proper weighting formula?