I Hired for Ticket Volume. Now I'm Rebuilding for Judgment.
Three years ago, I walked into a performance review with one of my team leads and we chatted about one of their best support agents. Numbers don’t lie. She was hitting 8 tickets per hour. Quality scores in the 90s. Absence rate solid. By every metric I had built the team around, she was a star.
Five minutes in, she told me she was leaving.
Her reason stuck with me: “I close tickets. I don’t solve problems.”
At the time, I didn’t really understand. I was too busy backfilling. Too focused on how we’d hit our deflection targets without her. But looking back now, she was pointing at something I’d built into the entire organization without realizing it. A machine optimized for speed.
That was the signal. I just ignored it.
When I started building support teams, throughput was the honest metric. It had to be. You’re managing a P&L. You need to know what it costs to handle a ticket. You need to forecast hiring based on volume. Tickets per hour, average handle time, cost per resolution—these aren’t arbitrary. They’re how you survive.
But then something shifted. Not in support, but in what support does. When most customer problems were password resets and billing questions, AHT mattered. The fastest agent won. The person who could navigate Zendesk with one hand and type with the other would always be the best performer.
Then AI showed up. Then automation. Then the deflection movement. Suddenly, the fastest agent is handling the hard stuff that survived the first pass. The complicated edge cases. The customers who are already frustrated. The problems that require pattern-breaking, context, judgment.
Speed became a liability.
I didn’t rebuild immediately. Most leaders don’t. We layer the new metrics on top of the old ones and hope it works out. We tell people: “Hit your ticket targets AND improve your judgment.” What we’re really saying is: “Keep doing what you were hired for, but also do the thing that contradicts it.”
People feel that contradiction. They optimize for what gets measured. So they optimize for closing tickets, even the ones that need more time. Then we wonder why repeat contacts are rising.
The rebuild had to be from the ground up.
The first thing I changed was how we talked about performance in 1-on-1s. I stopped leading with numbers. Started asking: “Walk me through a ticket you weren’t sure about last week.” That one question opened up a different conversation. It moved us from “Did you hit your target?” to “How are you thinking?”
Some people leaned into it. Others got frustrated. They’d say, “I know what I’m doing—look at my metrics.” And I’d have to be honest: “Your metrics were built for a job that doesn’t exist anymore.”
That’s the hard part of rebuilding. You’re asking people to be better at something that, by old measures, makes them worse. An agent who spends fifteen minutes on a difficult ticket instead of closing three faster ones drops their AHT. On a spreadsheet, that looks like decline. In reality, it’s judgment.
I had to rebuild the scorecard. Started tracking “first contact precision”—did this ticket actually get solved, or just closed? Started looking at which agents were catching complex issues and routing them correctly instead of oversimplifying for closure. Started measuring what happened after the ticket closed.
The rankings changed completely. Some people stayed at the top. Others climbed. A few realized the shift wasn’t for them and moved on.
One of my highest performers under the old system ended up middle of the pack under the new one. She was fast, reliable, and terrified to deviate. I had built a person for throughput. Trying to shift her into judgment mode felt like asking a sprinter to run distance. I recommended a transition into operations. Better role, better fit. She was relieved.
On the flip side, I had someone who’d always been “fine”—solid metrics, nothing special. Nothing broken. Once I started rewarding people who asked hard questions in real time, who consulted other team members instead of closing quickly, who flagged patterns to product—she lit up. She was actually great at solving critical issues. She just didn’t know it because the system had never asked her to.
Rebuilding also meant changing how I hire. For years, I’d bring in people with high-velocity backgrounds. Retail supervisors. Call center veterans. People with proven ability to move volume. Now I’m looking for something different.
I want curiosity. I want someone who reads the ticket and sees a question, not a task. I want people who’ve actually broken down why a customer was frustrated instead of just solving what they asked for. I want to hire problem-finders, not problem-closers.
The interview changed. I’m asking: “Tell me about a time you realized the obvious solution wasn’t the real solution.” I’m listening for how they describe the thinking process, not the outcome.
This takes longer to hire for. You can’t assess it in fifteen minutes. But it matters because you’re no longer hiring for an assembly line. You’re hiring for judgment.
The philosophical part—the part I didn’t expect—is that this rebuild changed what leadership means.
When you’re optimizing for throughput, leadership is about systems and discipline. Keep the queue moving. Remove blockers. Enforce standards. I was good at that. I built tight SLAs and knew every reason why a ticket sat in queue too long.
But when you’re optimizing for judgment, leadership is about creating space for thinking. It’s about asking better questions in coaching instead of pointing out where they deviated from the script. It’s about protecting people who challenge the easy answer. It’s about saying “I don’t know either, let’s figure it out together” instead of defaulting to policy.
That’s harder. Systems are predictable. Judgment is not. You can’t scale it as easily. You can’t forecast it as tightly.
But you also can’t rebuild your entire support org every two years because people burned out closing tickets they knew weren’t actually solved.
The reality is both things are true. You still need throughput. You still need metrics. You still need to manage a P&L. I’m not preaching some utopian “let’s just focus on quality and happiness” fairy tale. That’s how you lose business.
But the mix has to shift. The center of gravity has to move away from speed and toward depth.
If you’re still hiring for velocity, rewarding fast closures, and measuring success primarily through AHT, you’re building an organization for a problem that automation is solving. You’re asking people to compete with machines at what machines do best. And you’re missing the actual value—which is judgment, context, knowing when to bend the rule, seeing the pattern everyone else missed.
I hired for ticket volume because that’s what we needed then. Now I’m rebuilding for judgment because that’s what we need now.
The person who left taught me that. I just needed three years to listen.