A Green SLA Does Not Mean a Happy Client

Every operations leader knows the line: what gets measured gets managed. It works. Measure SLA compliance and the dashboard turns green within a quarter. The problem starts when you look at what actually produced the green.

"When a measure becomes a target, it ceases to be a good measure."

Charles Goodhart, as popularized by Goodhart's law

Tell a support team their job is to hit the SLA, and the team will hit the SLA. How they hit it is a different story.

Four ways teams game the clock

None of this is malicious. Each behavior is a rational response to the incentive you installed.

Empty comments that reset the timer. A note that says "still investigating" adds zero value to the resolution. Its only purpose is keeping the ticket from going red.

Premature closures. The ticket closes, the client complains, a new ticket opens. The dashboard counts two resolutions. The client experienced one problem that took twice as long.

Ticket splitting. A complex problem becomes five tickets, each with its own comfortable clock. Compliance looks great. The actual issue took a week.

Cherry-picking. Agents grab the easy tickets first because those protect the personal metric. The urgent, messy ones wait. Which means your SLA policy is quietly deprioritizing exactly the work that matters most to clients.

The metric was never the job

The management mistake is singular: making the team's job about hitting the SLA. The real job is solving the client's problem. The SLA is a result, not a goal. When you aim at the client, compliance follows and it is sustainable. When you aim at the number, the client becomes an obstacle between your team and a green dashboard.

You can see which mode an operation is in by checking two numbers nobody puts on the wall: reopen rate and time to actual resolution, measured from the client's first report to the moment the problem stopped existing. If those two diverge from your SLA compliance, the compliance is theater.

Why repetitive tickets make it worse

Clock pressure concentrates where volume is. Password resets, MFA lockouts, VPN drops, access requests. The tickets are simple, but there are dozens of them every morning, and each one carries its own timer. Your agents are not gaming the clock on the interesting work. They are gaming it on the queue that never stops refilling.

This is the part of the queue where the incentive problem dissolves if a human never touches the ticket at all. A system that resolves a password reset end to end has no personal metric to protect. It does not write timer comments. It does not split tickets. It does not pick the easy ones first, because it takes all of them at once. The clock simply stops being something anyone needs to manage.

And the humans who used to carry that queue get their attention back for the tickets where judgment matters, with no incentive to rush them, because the volume pressure is gone.

What to do this quarter

Start measuring reopen rate and true time to resolution alongside SLA compliance. Do not announce targets for them yet. Just watch the gap between what the dashboard says and what those numbers say.

Then look at where your timers concentrate. If most of your clock pressure comes from repetitive L1 volume, that pressure is removable. Not by asking the team to behave differently under the same incentive, but by taking that part of the queue away from the clock entirely.

The SLA is not the enemy. Making it the only target is.