CIOs face IT knowledge loss as baby boomers retire

Critical IT knowledge often walks out the door when experienced employees retire.

As baby boomers reach retirement age, CIOs face a growing risk of losing employees who keep critical systems running. Organizations will lose any undocumented expertise, judgment or historical context that those employees carry. Similar to technical debt or cybersecurity vulnerabilities, knowledge loss can quietly accumulate until it causes an outage or failed migration.

In this Q&A, Joe Locandro, CIO of Rimini Street, explains why CIOs should treat knowledge loss as an enterprise risk and how to best address it.

Editor's note: The following transcript was edited for brevity and clarity.

What risks do CIOs face when experienced employees near retirement?

Joe Locandro: When experienced employees leave the organization, the company doesn't just lose the capacity of the individuals themselves, but it loses a lot of knowledge, context and nuances behind the systems. I put knowledge loss on the same risk agenda as tech debt, cybersecurity or business continuity. The problem is it doesn't always show up immediately when companies downsize or when people leave or retire. It slowly accumulates until something happens.

An undocumented expert -- someone whose knowledge exists mostly in their head -- is just as much of a business risk as an unsupported application. The systems might appear stable on paper, but the real operational knowledge lies with just a few people who know all the customizations, integrations, workarounds and upstream and downstream effects. That's why it's important to define that as another risk.

Which systems or parts of the IT environment are most vulnerable when experienced employees leave?

Some of these systems were developed in the '90s or early 2000s, and only a few people knew how they ran.

Locandro: It's the mission-critical systems, and you must identify them. I did a technical debt assessment in my last job, and we looked at all the systems that would have a high customer or manufacturing impact if they went down. We found many single points of failure because some of these systems were developed in the '90s or early 2000s, and only a few people knew how they ran. Then, in your risk register, you say these systems must be either renewed or supported, etc. That's how the journey starts.

How can CIOs identify where they're most at risk for losing critical knowledge?

Locandro: There's a lot of documentation in playbooks and run books, but what's missing is the tacit knowledge. Experienced employees know why things were done -- why they did the workaround. Even when things go wrong in support, they know what to avoid. It's not just about documenting a process. It's much harder to document judgment. That tacit knowledge, which includes the trade-offs that they made, is the real challenge for CIOs to get a handle on.

Experienced people often say things like, "I've seen that happen three times in 10 years -- I know what to do." That's tacit experience versus what's in the document, which may or may not be up to date.

What does capturing tacit knowledge look like in practice?

Locandro: If you study anthropology, you'll see that tacit knowledge has been handed down through storytelling. Storytelling is the best way for people to transfer wisdom to future generations. The idea is not to start documenting but to hold workshops.

Storytelling is the best way for people to transfer wisdom to future generations.

Do your knowledge risk map as you do your technical risk map, look at which processes are most critical and consider who the owner is and who supports them. Then start a process well before people retire, before you downsize or before you do mergers and acquisitions, and ask them to tell stories. Ask them to look at some documentation or give experiences of when things have gone wrong, what they've done and the logic behind it.

I started looking at recording these storytelling sessions in my last job. Now, with the advent of AI, you can get these into transcripts, record them and supplement them with formal documents. So, you have two types of documentation: the formal documentation -- system operating procedures, playbooks, etc. -- and then you have the workshops about how things work, how they connect downstream, and the dos and don'ts.

Should CIOs only record sessions with people who are soon to leave the organization, or is this something you do regularly as you go?

Locandro: It should be something you do regularly once you've defined your critical single points of failure. However, most organizations don't do it because CIOs and technical people are focused on hardware, software and support -- they do it only when there's a need, rather than as an ongoing discipline. That needs to change because we know demographically, with the baby boomers, the aging workforce is upon us, and a lot of these systems are being supported by people in their 50s and 60s.

I mapped the age profile against the people supporting [my critical systems] and found that within five years, many were eligible to retire.

We also know that technology requires transformation, digitization and moving to platforms. That means you need to start planning now. After I identified my critical systems, I mapped the age profile against the people supporting them and found that within five years, many were eligible to retire. Then I went to the board and told them that we have an institutional knowledge risk: Even if we want to migrate to a new platform, it would take three to five years, and many of the people supporting the existing one are due to retire.

We need a plan to mitigate that risk. When you do a five- or six-year plan, it's quite scary how many people have that institutional knowledge.

Knowledge hoarding is a real thing in IT. How can CIOs encourage employees to share what they know?

We had the discussions with them and said we would like to work with you to start documenting all that you know.

Locandro: We had discussions with these employees and told them that we knew they wanted to change their lifestyle, retire and spend time with their grandkids. So, two or three years before that occurred, we had the discussions with them and said we would like to work with you to start documenting all that you know, so we can relieve you and you can go to your next chapter of life to retire. We don't want to bug you two or three days a week or ask you to be on call, because it's a very unique system and only you and two others know about it. Make them feel comfortable that you're supporting them in their next chapter of life.

However, some don't aspire to retire and think they'll be there forever. And there's another way to get that same knowledge: Go through all the meeting transcripts, service tickets or project histories, and this is where AI can become a librarian. You could say to the AI, "This is our plant manufacturing system. I want you to go through all the tickets from the last 10 years and summarize all the faults and fixes. I want to go through all the projects for upgrades and enhancements and summarize and document what occurred and the logic of why they were done."

You can do a whole lot of mass collection that wasn't possible five years ago, and you can start building up a knowledge repository from the fragmented sources across the organization.

Tim Murphy is a site editor and writer for the IT Strategy team at TechTarget.