Choosing to share metrics externally is a big step.

It's not just a moment-in-time decision; it's ongoing.

For the foreseeable future you'll get "In Dec of 2019 you reported X was Y. How has it changed since then?" kinds of questions.

If you don't answer those questions (or share the updated data proactively) people will assume that it's because the metrics look bad.

This is true for each and every individual metric you choose to share.

If you share the same metric multiple times on a given implied cadence, it becomes super-linearly hard to avoid breaking the precedent of sharing.

When you stop sharing, people will assume for sure that it's because the numbers aren't impressive, in proportion to how established the precedent for sharing is.

A counter-intuitive best practice for sharing a new metric: don't share it again in the time period after it was first shared, to avoid setting a precedent.

Also, share it only at random, hard-to-predict intervals, making it less obvious when you stop sharing it, and less likely to get "the metrics must be bad" suppositions.

(Of course, if you stop sharing it for good, at some point that will be obvious, no matter how random the interval... it will just take longer for people to notice.)