Product development performance is largely evaluated on product velocity, which creates an incentive to push new code as quickly as possible. Meanwhile, SRE performance is (unsurprisingly) evaluated based upon reliability of a service, which implies an incentive to push back against a high rate of change.
