The following post was written by Ralph Laura, CIO, HP Enterprise Group, Global Sales, HP Labs and originally published in the HP Infrastructure Insights blog.
My youngest daughter recently “refactored” her bedroom. I would say she cleaned it, but this was far more than a cleaning, even of the “seasonal” variety. As she moved from young girl to teenager she felt it was time to perform a wardrobe, desk, accessories and treasured artifacts review, and restructure her belongings and bedroom to better match her evolving personality and preferences. She was shocked how much room there was (and how big the donate or discard pile in the hall grew) as she went through this process.
Similarly, no IT organization sets out to have thousands of applications in their environment. Every app had a purpose - at some point, or it wouldn’t be in the datacenter. The go-to solution for IT to address a specific business need or deliver a crucial outcome is often to purchase or develop a new application. But without proper stewardship, an ecosystem can become cluttered with the accumulated layers of prior behaviors and preferences. I’ve often heard this referred to as ‘IT debt’ or the ‘IT tax’ in an organization.
Anyone who has spent time in an IT department has seen how application, device, storage, etc. sprawl can drain an operations and support budget, so why do we continue to repeat the pattern over ...
[Unchecked PC proliferation, followed by PC lifecycle management solution deployments]
and over ...
[Unchecked mobile phone proliferation, followed by MDM and later EDM solution deployments]
and over ...
[Unchecked use of non-enterprise cloud services, followed by deployment of cloud access security brokers]
The common thread in most of these is that at the initial scale these solutions all seemed harmlessly cute and furry in small numbers (like our friends the Tribbles), but as they multiply they can overwhelm the resources of the enterprise (in both the general and the starship context!). They can quite literally consume all available space and resources, leaving you with the unenviable choice of slashing core budgetary line items to continue to feed the unmanaged multitudes of PCs, Mobile Devices, Apps, etc. or play the ogre and start culling the herd, forever to be known as the grim reaper of all things cute and fuzzy!
So, how do you avoid this inglorious fate? The first step is to size the problem. This is not a cursory, or even detailed look through the enterprise asset database. This involves cleaning in the corners in the back of the legacy platforms you thought had been retired years ago, rifling through the attic of one-off solutions that are still alive and well on a tower under a desk somewhere, and most importantly, looking to the vastness of the cloud. The best way to get a real handle on use is to go to where the users are (the edge) and see what they are doing (things they often neglect to inform you of). There are a number of tools in the Cloud Access Security Broker (CASB) space that can help in this area, and a whole new class of endpoint solutions that excel at finding what you didn’t know to look for.
The next step is to get a real handle on costs. Many of these small cloud or app platforms seem negligible in cost, until you add them across hundreds, thousands, or tens of thousands of devices and users. I’ve long been a proponent of a services-based costing view in IT, and helped to found the Technology Business Management (TBM) Council several years ago - a tremendous resource in this area (tbmcouncil.org).
Once you’ve established a view of current apps (not a one time inventory, rather a proactive and ongoing process) and mapped them to the operating model of your company in a TBM-based platform, patterns and trends begin to emerge. Shining light on a major source of cost or resource consumption may identify some low-hanging fruit. Maybe you can pull the plug on a few orphaned apps, regaining compute power and other resources. In the cloud space you may find various departments selecting diverse solutions for the same basic capability, like file sharing, research, or lead generation. If IT can negotiate enterprise-wide contracts and reduce costs, the business becomes much more amenable to standardizing on a platform.
So now you’ve counted the Tribbles and tallied up their dinner tab. Nice job, but from here the work gets more delicate. It may be necessary to tell a valued business stakeholder that their pet application is obsolete or inefficient. But now you’ll be armed with solid data showing what it’s really costing to maintain, and what you may be forced to sacrifice if they can’t agree to an alternative.
In every company, legacy and monolithic/fragmented apps are common. But just because an application has been around for a while doesn’t mean it must go. I like to use the reframing device that some apps are more “Heirloom” vs “Legacy/Depricated”. Often older, highly customized applications are still delivering strong value and, if properly maintained, can continue to do so.
But you want to carefully inspect overlapping application functionality and cases where multiple apps are supporting the same business processes. Stakeholders may say customization by region, or for varying products or go-to-market models is critical, and in some cases it may truly be. But IT can help the business envision “IT debt” in terms of ROI and/or risk: What value is the app creating for the cost it generates? Does the app present a risk in terms of future ability to transact business, or business continuity or agility in the face of future disruption? When costs and risk are presented in such terms the stakeholder may find consolidation or migration much more palatable.
Another route to simplification is to ‘declutter’ the application landscape by presenting a thoughtful environmental tool like HP’s Grommet framework, which can harness the functionality of a group of applications and present a single interface with which the user interacts. Since it’s not necessary for the user to interact with the applications themselves, IT can plug in and unplug apps in background. You can select functions and/or layers of capabilities to move to SaaS or private/hybrid cloud – and since the user continues to operate in that familiar interface the transition is more seamless.
We’d all prefer to focus our program dollars on change the business projects we know are critical to a modern datacenter, but if you’re not doing basic ‘garbage collection’ in the background you may be stranding resources or leaving money on the table. As with any major IT transformation, it’s a good idea to begin addressing application sprawl on a targeted scale. Start small, in one business process area or with one trusted stakeholder, show clear success in the stewardship of IT resources, and gain credibility and support. Then you’ll likely find other areas of the business beating a path to your door.
Facing an IT portfolio with a plethora of hungry applications can be daunting. But by applying some vigilance and stewardship to the environment it’s possible to develop a strategy that gets you out of the trouble with Tribbles. And framing stakeholder discussions in business terms backed by solid data can ‘demystify’ IT, help IT deliver clear business value, and make for a much stronger partnership between IT and business. In this way you may find that you both….wait for it…”Live long and prosper!” J
About the Author: Ralph Loura is a key leader for the newly formed Hewlett Packard Enterprise, serving as CIO for the Enterprise Group and Hewlett Packard Labs. Ralph’s objective is to help shape go-to-market strategies through world-class processes, tools, and data that identify opportunities for HP’s Enterprise Group. He believes all IT functions should serve as value creators rather than cost centers, and that the path to value comes from business outcomes tied to user experience.
For additional insights on Ralph Loura’s take on this topic, check out the webinar titled Application transformation: The difference between Heirloom and Legacy IT