It is the best and worst part of a meeting. After everything from sports analogies, pain and gain, speeds and feeds, and time on the whiteboard, the CIO might turn to you at minute 42 of a 60-minute presentation and ask, “what do you think we should do?”
White knuckle moment? Victory lap? Time to lob another question? Strangely, what I find myself saying these days, is that IT organizations should resolve their unaddressed debt.
To be even more difficult, I point out that not only are there multiple types of debt, not all IT debt is the same. Furthermore, what used to be the most critical type of debt to address may no longer be your top priority.
In most IT organizations, paying down technical debt takes up all the airtime. Fifteen years ago, with limited alternatives to running everything in your data center, if you couldn’t properly manage your technical debt, you would put the entire business at risk.
Whether it was upgrading components, solving interop issues, replacing slow spinning disks, or even mundane tasks such as firmware upgrades, technical debt was the bane of every IT decision maker’s existence. And for many IT organizations, this is still the case. KSR (Keeping Stuff Running) is job number one. High availability and performance are dependent on low levels of technical debt.
Technical debt can prevent even world-class IT organizations from succeeding. As recently as 2015, you could find statements in the CIO Journal on wsj.com noting that, “more CIOs are turning their attention to technical debt, they want to identify and categorize it, quantify the cost of it, and estimate the cost and benefit of paying it down.”
So why would I recommend deprioritizing technical debt?
In 2020, the options available to resolve technical debt are very different than they were in the early 2000s, or even five years ago. Most infrastructure can now be defined and managed through software. Upgrades, migrations, lifecycle management, and even rolling deployments can be automated and centrally managed.
When it comes to IT infrastructure, does paying down technical debt always provide the highest return to the organization or business? Are there other types of debt that are more critical, both from an economic and a technology perspective?
Thought experiment: you are the VP Infrastructure at a mid-sized insurance company, and you decide to move all your legacy workloads overnight to a third party Infrastructure as a Service provider. This is a thought experiment, so magically the transition from your data center to a new IaaS backend goes flawlessly over a single weekend.
On Monday morning, when your team returns to work, what percentage of their time will still be sacrificed to resolving infrastructure technical debt?
Very little is a fair assumption. It is more likely that your sys admins and architects’ efforts will be focused on other priority areas like threat management, developer productivity, and cost.
This thought experiment crudely represents the opportunity that many organizations see today. If their infrastructure technical debt was entirely, or even mostly, resolved through next-generation offerings such as VCF, VMC , or public cloud IaaS services, what would the priorities be going forward?
Which brings us back to the original question, “what do you think we should do?”
IT resources are precious, whether you are in the public cloud, the data center, or somewhere in between. Therefore, it is essential to know how your efforts uniquely contribute to your organization’s success. If your team could double the time spent on critical application outcomes and radically efficient service delivery, what would the impact be on morale, your mission, and the organization at large?
My suggestion is to look past your technical debt and start assessing the current state of your other IT debt categories. Not only will this support your future success, but it will also help you to see how to best address your technical debt with the highest rate return on your time and money. More on that later in ‘Cloud Is’ Part Eight.
For now, what are the other debts to pay? The first is policy debt, which represents the non-infrastructure aspects of your IT and cloud service management overhead. Policy debt will stay with you, even if you have a real-world thought experiment where technical debt disappears over a weekend.
Next week: ‘Cloud Is’ – Part Five – Policy Debt
The full ‘Cloud Is’ series is available here: https://blogs.vmware.com/management/author/rquerin