Track impacts of required Verify approvals

Problem

With Enforce domain expert approval in critical CI a... (#393864 - closed) we introduced required approvals for Verify backend and frontend changes.

We don't have a way to monitor the impacts of this change. As a Verify leader I want to track how required approvals affect the engineering productivity and quality of the Verify code.

Ideas

• Did our mean time to merge or MR Rates decrease? (short term)
• MTTM or MR Rates for MRs authored by others and approved by Verify (short term)
• Did we prevent incidents due to this extra approval process? (mid-long term)
• Rate of bugs and maintenance issues created overtime (mid-long term)
• Do we need a satisfaction survey and get more data from engineers? (short term)

Details

Today we have http://app.periscopedata.com/app/gitlab/782405/Verify-Group-MR-Metrics and we deployed the first required approvals early April.

Solution

• introduce 2 new labels to be added when a MR is approved which required Verify approval: verify-reviewimpacted and verify-reviewnot impacted
• update handbook for the team to apply the labels.
• automate message on MR when a Verify approval is applied and the label is missing, to remind approvers to add the right label.