I don't know if 75+ instances over 18 months is large enough to count as quantitative data, but we've never lost a large instance (they've had temporary failure

I don't know if 75+ instances over 18 months is large enough to count as quantitative data, but we've never lost a large instance (they've had temporary failures). All four of the permanent vanishing-acts were smalls.