Fixed One Thing, Broke Another

Why this matters

You finish a repair, and now something that worked before is broken. It feels like bad luck, but it almost never is. A new fault that appears right after your work is connected to your work most of the time. The honest move is to treat your own repair as the prime suspect, not the last resort. Owning that gets the second problem solved fast and keeps the customer's trust.

Start here: assume you are involved

Timing is the clue. A symptom that began immediately after your visit is tied to your visit until proven otherwise. There are three ways your work can produce a new fault, and you check them in order.

Path 1: you disturbed something during access

The most common cause. Getting to the repair meant moving, disconnecting, or leaning on other things, and one of them did not go back right.

• If you disconnected anything to reach the work, confirm every one of those connections is fully remade. A control wire left off, a hose not reseated, a connector half-clicked, a fastener not retorqued.
• If you moved or flexed adjacent parts, a marginal connection you disturbed may have finally let go. A wire that was barely holding, a brittle fitting, a corroded terminal. Re-check anything your hands or tools touched.
• If you removed a panel, guard, or cover, confirm it is back in place. Some systems will not run with an interlock open, and the "new fault" is simply a guard not fully seated.

Retrace your access path physically. Touch every connection you broke and every part you moved.

Path 2: your repair exposed a pre-existing weakness

Sometimes your work was clean, but fixing one thing put normal load back onto a second thing that was already failing quietly.

• If the now-broken part is downstream of your repair, it may have been coasting on a degraded system. Once you restored full operation, the real load returned and the weak part quit. It was going to fail anyway; your fix just revealed it.
• If restoring one function increased demand elsewhere, a marginal component that survived under reduced load can fail under full load. The fault was latent, not caused by you, but exposed by you.

This case is not your error, and you should say so plainly to the customer - while still fixing it. "The repair worked, and it revealed a second worn part that was hidden until the system ran at full output."

Path 3: a setting or configuration changed

• If you changed a setting, jumper, or program as part of the repair, confirm it did not affect a shared function. One control often governs more than one behavior.
• If you reset or power-cycled the system, a setting that was held in volatile memory may have reverted to a default that breaks something else. Check configuration after any reset.

Telling caused-by-me apart from exposed-by-me

The distinction matters for how you handle it:

• Caused by my work (a disturbed connection, a missed step): your responsibility, fix it at no surprise to the customer.
• Exposed by my work (a hidden weak part that full load finally killed): a legitimate second problem, communicated honestly, handled as additional work.

Do not hide behind "exposed" when it was really "caused," and do not eat the cost of a genuinely separate failure as if it were your mistake. Be straight about which it is.

Prevent the next one

Before you button up any job, do a closing sweep: every connection you broke is remade and torqued, every panel is back, every setting is confirmed, and the whole system - not just your repair - runs through its normal cycle while you watch. The fault you catch on your closing sweep is not a callback.

References

• Trade-standard repair and closeout practice
• See related: Callback: The Fix Didn't Hold
• See related: Customer Says It's Worse After Service
• See related: Torque: Why It Matters and Getting It Right