This would depend on the usage metrics that are in decline and the product launching.
I would start with investigating why there is a drop in usage, as there are several scenarios -
1. if low usage is overall and due to seasonality, I would generally hold off on product launches.
2. if some of the usage metrics are down, and some are up, then it depends on the product itself -
a. If the new product is more relevant to the ones that are up, no reason not to go forward.
b. If the new product is more relevant to the declined metrics, then we need to understand if it's a steady decilne (trend seen overtime) and the new product might assist with it. Or, perhaps it's a sudden decline, and I would hold off with product launch until I have an understanding of what is causing this sudden decline.
Of course, if any publich commitments were made to the launch of this product, it's also something we need to take into consideration, since changing a public publish date is problematic for user trust.
