In order to streamline management, standardise toolsets and ultimately reduce costs, many companies are adopting the concept of software profiles in their software deployment strategies. These implementations are often supported by the deployment of Identity and Access management tools which control access and supplement business profiles.
The concept of profiles makes sense. The basic premise is to standardise technology and software deployments based on the tasks that groups of workers perform. So for instance the applications deployed in finance may be very different to those deployed in a call centre, although granted there may be some overlap which form part of the base profile.
From an IT departments perspective, and more particularly the software deployment team, this method of deploying infrastructure is a relatively simple with only a requisite need to manage exceptions thereafter. At face value this is far more efficient and thus cost effective. Furthermore, enterprise agreements typically entitle and encourage the proliferation of this software, now there’s a surprise. Where these often fall down however is non EA type licensing and department specific applications.
But is this approach actually saving money? Is there a hidden elephant in the room?
Unfortunately, I would argue that this approach is actually quite antiquated and to a large degree completely negates the concept of on-demand computing completely.
The elephant is software waste and the elephant is quite large. Software or user profiling is actually proliferating significant additional and unwanted costs in the form of software waste.
One size does not fit all and as a result, much of the software deployed through profile type deployments remains unused. Users get it for the sake of the profile, not because they are actually using it. This software waste not only attracts initial license costs but ongoing maintenance in the process. And this waste is significant, adding up to many millions of dollars per annum in unwanted costs in any reasonably sized organisation. In fact independent research commissioned by 1E indicated between $40 and $60 of unused licensed software per PC on average.
So if software profiling is costing money are there any alternatives? I would not be writing this blog if there wasn’t.
The solution is twofold, metering and user self-service.
Metering your applications will allow you to identify applications not utilised which can then be harvested for re-use or renegotiation with your vendors. What is important here is to have the ability to meter everything, not just specific vendors or specific applications! What you don’t know, you don’t know. That number crunching application in finance or that CAD application in engineering could be very costly and removing just a small percentage of these rogue deployments could make a significant impact on the bottom line. If you are going to meter, make sure you have the tooling to do it properly without impacting your infrastructure.
Secondly, you need to empower your users to take control of their desktop computing environment. This should be simple, intuitive, user friendly and accessible. User self-service ticks all these boxes and exposes the end user to ICT in a very different manner. You will be amazed what the impact of giving your users their own internal “Amazon.com” will have on their perceptions of IT. They will love you. If they need something they should be able to shop for it and it should get deployed in an automated fashion, no fuss with appropriate workflow and approvals.
The moral of all of this, particularly if you have a budget to maintain, is to scratch below the surface in terms of how your software is getting deployed. Are you shooting yourself in the foot when it comes to your next EA negotiation, maintenance renewal or even worse, vendor audit!
Author: Tim James, Director sustainableIT