Perhaps the most difficult phase of product development is sunsetting products. If a product has not gained traction with users or it's not aligned with company goals, the only reasonable thing to do is to first freeze its development, then stop accepting new users and finally disabling it altogether. (Un)fortunately even an unsuccessful product could (and usually does) have a base of loyal users to whom the product is important nonetheless. Quite often, the most vocal opposition to the sunsetting of a product, will be your colleagues which are probably using the product the longest and have invested the most in it. In such cases it takes a strong product manager to hold on decision and explain in detail to all stakeholders the rationale behind the decision, and even more importantly, what will the organization gain with more focus and freed resources. A good if a bit user unfriendly way of checking loyalty of users is to just disable the product for a few days. If no-one complains you get further validation that your product isn't loved by anybody and if they do complain, you can estimate the fall-out and either turn back on service or comfort the users in some other ways.
I must congratulate Google on handling of sunsetting of its Reader. While painful to many of us, it was the only right thing to do. Instead of putting it on cold ice (as Google usually does with its products) they announce their decision well in advance so that decent alternatives could pop up.