Additionally, a lack of data in QA likely contributed to the fact that the slightly different metric was not even set up for AoC and VoC. If we had a better understanding of the code and the metrics with some more data, this issue might've been caught earlier in the ticket lifecycle.