Things move a lot faster in library-land than people outside the profession would ever imagine - approaches, trends, philosophies and pedagogies all shift on an on-going basis. For this reason, the fact that something didn't work 2 or 3 years ago is really not a good enough reason not to try it now (and by the same token, the fact that something DID work 3 years ago isn't enough of a reason to keep doing it - we have to make sure it's still working in today's landscape).
This happens a lot though - someone new comes into an organisation and says 'why don't we try such and such?' and the reply is 'oh we already did that; it didn't work' and that's the end of it. In effect, a policy has been built off the back of one experience - and that experience may not be representative anymore, because things change, and people change.
This is particularly true in the web 2.0 landscape, where individuals' attitudes to interacting with organisations and businesses changes all the time. A Library may run a trial and the conclusion 'our users don't want to be friends with a Library on Facebook' emerges. If this trial took place in 2012 then it is entirely valid; don't waste your time and effort on a Facebook presence. If the trial happened in 2009, it's almost entirely without worth! That is SO long ago as to need re-visiting before a decision can be made on whether or not Facebook is a good idea - web 2.0 years are like dog years, so a 2009 Facebook study is the equivalent of a 1990 Library Management System study. :)
So, if you come across something that has already been tried, and you think the landscape has shifted sufficiently to try it again, don't take no for an answer! It may be that it doesn't work this time either, or it might be a huge triumph - either way, your Library's policy will be based on something current, and will be more likely to reflect the needs of your users...
p.s I was in a Lean methodology training session the other day, which is what inspired me to finish this post which has lain in my drafts folder for a couple of months. It turns out a lot of Lean principles are things I've been thinking about for a while, including the business of not just doing things one way because they've always been done that way, and not trying anything new because it has been tried once before in the distant past. Lean puts it in terms of the five whys - asking why (or more likely, 'yeah, but WHY though?') enough times to actually get to the root cause of something. Heidi Fraser-Krauss who led the session gave an example of a hospital who asked their staff to sign into a book when they rode their bikes into work. No one knew why, it had apparently ever been thus. The bike-book went back as far as the 40s and was, it turned out, something to do with rationing during the war... So it just goes to show, some processes need a quick currency-check to see if they're still needed. Eliminate waste.
We're being asked to take on so many new functions as part of the changing role of the Information Professional - if we don't make sure we also lose anything non-essential, we'll eventually run out of steam...