Situation
Between 2017 and 2019, I owned L1–L3 application operations across a 150-system production portfolio at a Tier-1 carrier — every reactive ticket on every shift came out of the same team's hours. One of the senior L2 engineers on the team was stuck in the classic ops trap: highest ticket-close rate, the on-call everyone wanted, no upward path. The promotion case kept coming back the same way: “high output, but no scope growth, no architectural footprint.”
Task
Project Manager for Operations, accountable for 14 direct and ~50 indirect reports and for the career arcs of the senior engineers on the team. Growing this one engineer wasn't optional — they were the bar I needed the rest of the team to reach.
Action
I stopped treating ticket-close rate as the headline metric in our 1:1s. I asked them to pick one recurring incident from their own queue — one they'd seen three or more times that quarter — and own the permanent fix end-to-end instead of closing the next instance of it. I unblocked the portfolio-wide access they needed to ship the fix (a thing they'd previously had to escalate to me to get). I covered the on-call rotation they'd normally have taken so the calendar space existed to do the deeper work. Then I made the next engineer on the team do the same thing — and the next — until “find your repeat ticket and kill it” became the team's actual practice, not a slogan in a deck.
Result
Multiple permanent fixes shipped in the first quarter, and the practice compounded across the team. The portfolio-level outcome was the 60% stability improvement over three years — but the part I'm prouder of is that the change compounded because the senior engineer went first. The principal-level lesson: people don't grow into the work you ask them to do. They grow into the work you stop asking them to do.