Raise what counts as acceptable before raising what counts as impressive.

Working on your best output while leaving your acceptable output where it is produces a wide gap and an unreliable body of work. Raising the minimum is less glamorous and affects far more of what you actually produce, since most of what you make will sit near the acceptable line rather than at the top.

Define what you will no longer let pass, and make it specific. Then hold that line for a month on everything, including the work nobody will examine. The overall standard rises considerably more than it would from another push at the top.