If your organization is big enough, could be worthed to use CVMFS
Now, what is CVMFS.
CVMFS (Cern Virtual Machine - FileSystem) is one of the ways we deploy software inside CERN, given the sheer amount of computation unit we manage it made sense to develop a solution specific for our needs.
Basically, there is a server that talks HTTP and provides a file catalogue along with the files themselves.
The clients connect to the server and using FUSE populate a specific directory /cvmfs/ by default with whatever software is installed on the server.
You also get a tag system, a little bit like git, where you can jump between different tags.
Maybe for 10 machines is not worthed, but if you scale it on the order of 50/100 you will definitely get some advantage from it. If you scale beyond that is the only scalable solution I know of.
If you are interested, get in touch 
Let me leave you the link to the documentation: Welcome to CernVM-FS’s documentation! — CernVM-FS 2.9.4 documentation and the one of the code: https://github.com/cvmfs/cvmfs