For our customers, it is not a question if they should consider data centre optimization, but rather when and how. Since the “when” question is almost always related to business priorities and budget, I will focus on answering the “how”.
The most important area to have a look at is the IT Infrastructure environment itself. For example, how much do you pay monthly on HW maintenance fees to your storage subsystem vendor? Do you have any 3+ years old storage subsystems in your environment? Did the maintenance costs increase after the standard 3-4-5 year maintenance contract purchased with the hardware expired? Are you on extended HW support contract? If so, decommission the hardware immediately.
According to my experience, maintenance costs of storage subsystems in their 4th and 5th year could be higher compared purchasing new hardware, installing the new subsystem and migrating the data.
Good practice to manage data migrations is to add a storage virtualization layer. It allows you to treat underlying storage subsystems as a commodity.
Similar arguments apply for virtualization of host systems as well.
The goal of introducing virtualization to your environment is to standardize your environment from the management and data migration point of view. Standardization of the environment so that it consists of unified building blocks is the prerequisite for successful introduction of automation and provisioning to your environment.
Once your environment consists of standardized building blocks, you could start looking at task automation and automated provisioning, such as (re)deployment of OS images, standard software packages installations, patch management and so on.
By focusing on replacing your aged commodity storage and systems hardware, you could save substantial OPEX costs.