Agile is different, than other form of software development practice. When I say different, it's different not only the way we deliver software, communicate with user, build preparation but at same time we test in shorter cycle. SDET (Software development engineer in test) plays an important role in agile testing. SDET on a agile model works more than a role of a tester, its more of a role of customer advocate.
Testing metrics in agile is crucial, as agile team is least bound by process advocacy. Team should implement metrics in an agile environment which really adds value.
One of such metrics which we use in our project is "100% Done Criteria Vs. Partial Done Criteria".
This is a simple metrics, while says if UAT Users (Product Owner) fully satisfied with Done criteria or If user story 100% met it done criteria, or user accepted a user story partially.
You may raise question, without 100% done criteria been met, a user story can never go to production, yes, but some times unusual things may happen.
So this metrics is a checker during agile cycle, that all done criteria properly tested and accepted.
I will love to hear feedback from practitioners about this metrics (positives and negatives).