"We've carried this one a while. It's not giving us trouble. Let's push it to the next opportunity."

Heads nod around the table. The window's tight, the budget's tighter, and there are more jobs than anyone can realistically execute. On its own, the call feels sensible. Low risk. Reasonable.

Now run that same decision across dozens of assets, several units, and a few years. Nothing blows up right away, so the pattern goes quiet and becomes normal.

That's how most maintenance debt gets made. Not in one dramatic decision. In a thousand small, rational ones that each borrow a little against asset health and bet the future will be kinder than the past.

The obvious way to take on debt is to defer work.

Push a PM, an inspection, or a known repair and you're trading near-term time and money for future risk, right out in the open. Deferring now and then isn't the problem. That's reality. The problem is when deferral stops being the exception and turns into the habit.

The logic in the room sounds fine. We've never had a failure there. We inspected it last time and it was fine. We'll have a bigger window next outage. Sometimes all three are true. But every deferral nudges you one step further from the assumptions your maintenance strategy was built on. The interval that used to be conservative quietly turns aggressive. The line for taking action keeps moving.

Deferral is a loan. The interest shows up as a higher chance of failure, a worse outcome when it does fail, or a fatter repair bill when you finally deal with it.

Here's a quieter loan. Instead of deferring the whole job, you thin it.

To make the hours work, you trim inside the work order. Drop a couple of inspections from the PM. Skip part of the test. Cut back the disassembly or the cleaning. Do the main check and wave off the secondary one. The PM still exists. It still shows up at the same frequency. Your compliance report still reads green.

But you've quietly cut the actual work content by a third, maybe half. You're doing less maintenance while reporting the same compliance. That's drawing down the asset's equity and booking a debt the dashboard will never show you. When something fails later, it's easy to say "but we did the PM." On paper, sure. In practice, not really.

Operators and techs are resourceful. When the system's too slow or the plant has to keep running, they find a way. And some of those ways never touch your CMMS.

Temporary repairs that quietly become permanent. Bypasses and tweaks to get through the shift. Small fixes done on the fly and never logged. In the moment, that shadow work is a lifesaver. It keeps production up, skips the paperwork, and runs on hard-won local knowledge.

Stretch it over time and it builds debt two ways. The asset's real condition drifts away from what your records claim. And the next planner makes decisions on assumptions that are already wrong. You can't run precise, risk-based maintenance when a chunk of your history lives in people's heads instead of your system.

Another withdrawal. Accepting "good enough for now" on quality.

Under the clock, the temptations stack up. Skip the precision alignment or the balancing. Use the almost-right part because the exact one isn't on the shelf. Rush commissioning and the functional checks. Take the work without real QA. The asset comes back online. The outage lands on time. Looks like a win.

What you've usually done is shave the margin out of that asset. Bearings run hotter. Seals give out sooner. Vibration creeps up. The unit may ride out this campaign, but you've prepaid for a future failure. Every quality shortcut is one more withdrawal from the reliability account.

The last one hides in your data.

Close work orders with generic failure codes, a one-line comment, no cause, and no record of what actually got done, and you've robbed the next team of the context they need. When that asset comes up for planning again, the history doesn't tell the story. Was this a one-off or a chronic headache? A permanent fix or a band-aid? A design problem, an operating condition, or a gap in how it's maintained?

Without that, planners and engineers either reinvent the wheel or guess. Both roads build more debt, either by repeating half-fixes or by over-scoping because nobody trusts the record. Data debt doesn't just sit there. It multiplies every other kind by making it harder to see and harder to fix.

None of this is a demand for perfection. Real plants and real budgets force hard calls. You'll defer some work. You'll thin some scope. You'll accept the occasional shortcut when the alternative is worse.

Borrowing against asset health isn't the sin. Doing it unconsciously and untracked is, then acting shocked when the interest lands as failures, downtime, and capital nobody planned for.

Name these behaviors out loud as debt and the conversation changes. If we defer this, are we good with the debt we're taking on? If we thin this scope, what exactly are we cutting, and what risk rides along with it? How much shadow work is happening, and can we at least capture the story after the fact? Are we willing to pay the interest on this shortcut later, and do we know what that bill looks like?

In execution work, this is where we earn our keep. Not by stopping every deferral or shortcut, but by making the borrowing visible and deliberate, so the call lands on the record as a choice instead of slipping by as a habit nobody owns.

Maintenance debt almost never arrives in one big decision. It accrues a little at a time, in deferrals, thinned scope, shadow fixes, rushed quality, and silent records, each one reasonable on the day it gets made.

You'll keep borrowing. That's the job. The difference between a plant that manages its debt and one that gets blindsided by it comes down to this: one borrows on purpose and writes it down, the other borrows by reflex and forgets.

Next in the series, we head back to the complexity side and one big reason this borrowing stays invisible. When every job and every event gets forced through the same one-size-fits-all process, nobody's matching rigor to risk, and the debt just piles up in the blind spots.