The policy makes sense only because people are setting simple passwords. However, the systems I've seen also limit password length, which effectively encourages people to pick simple passwords suffixed by 1 or 2 digits.
Making people change their password is a bad practice from a different age, IMO. Such orgs these days are only safe due to proliferation of 2FA, IMO.
@jerry @sidd I tend to think that monitoring is your best tool. Multiple failed logins, geographic improbabilities, unexpected devices, etc. If you can force MFA challenges under such circumstances, you reduce much of the risk for day-to-day accounts. Use k-anonymity and/or breached password checks in the password change flow.
Then apply your more stringent standards to higher-risk accounts (privileged accounts, high-value targets, etc.) I think a targeted approach, coupled with good communication with end users, gets you so much closer to a regime everyone can support and live with.