Your support north star is probably an activity metric
A single guiding metric can align a whole team — but most support orgs pick one that measures their own motion instead of the customer's result. The case for steering by an outcome.
On the wall of one support floor, a single number glowed in meter-high digits: tickets closed today. It updated live. Reps could watch it climb as they worked, and by four in the afternoon a good day and a bad day were visible from across the room. It was, in every respect, a proper north star — one number, universally understood, impossible to ignore. It was also the wrong one.
The north-star idea comes from product teams: pick the single metric that best captures the value you create for customers, and align everyone behind moving it. Done well, it cuts through dashboard sprawl and stops every function optimizing its own corner. Support has adopted the ritual enthusiastically. It usually picks the wrong star, and it picks it the same way every time.
Activity metrics feel like north stars. That's the trap.
The numbers support reaches for first — tickets closed, average handle time, first response time, deflection rate — share a family resemblance. They measure the team's motion: how much you did, how fast, how cheaply. They move every day, respond instantly to effort, and sit under your direct control. That is exactly what makes them feel like north stars, and exactly why they fail as one.
The failure mode is Goodhart's law, and support demonstrates it better than most fields. Make deflection the number everyone watches and you will get more deflection — including the kind where a bot fires a confident non-answer and the customer quietly gives up. Even with self-service everywhere, Gartner found that
so a deflection star counts every contact it steered away from a human as a win and stays silent about the majority who never got their answer. Deflection dressed up as resolution is the purest example, but every activity metric has its own version: chase handle time and reps close fast, not well; chase first response and you get a cheerful "looking into it!" that resolves nothing.
Meanwhile the thing you actually care about can rot while the number glows green.
An activity dashboard cannot see that customer leave. It recorded a closed ticket and moved on.
An outcome metric measures the customer, not you
An activity metric measures how hard your team worked. An outcome metric measures whether the customer got what they came for. Only one of those is worth steering by.
The fix is to pick a star that lives on the customer's side of the glass: did they get what they came for, and how hard was it? The best-evidenced candidate is effort. When the old Corporate Executive Board — now Gartner — studied loyalty across tens of thousands of service interactions, the split was stark,
and the effort question turned out to be the sharpest predictor they had:
That is the whole case for an outcome north star in one line: the metric that best forecasts whether a customer stays measures their experience, not your throughput. Effort, not delight, is what moves loyalty.
The obvious objection is that satisfaction is already a customer-side metric — so why not make CSAT the star? Because it makes a poor one. Nationally it barely moves,
and what movement you do see is often your response rate shifting rather than your service. A north star has to tell a good week from a bad one. A metric that sits flat for a decade cannot.
Why the outcome star is worth the trouble
Outcome metrics are harder. Effort and honest resolution take surveys, repeat-contact tracking that follows the customer across ticket numbers, and a willingness to argue about what "resolved" really means. Activity metrics are sitting in your helpdesk already. Teams pick the easy star because it's there, not because it's right.
But the outcome star is the one that connects to money.
Support that reliably lowers effort and stops problems from coming back is retention work, and retention is where support earns its budget. No volume of tickets-closed points at that. A falling effort score does.
None of this means throwing the activity metrics out. Handle time still plans your capacity; first response still tells you whether the queue is on fire. They are instruments, not destinations — diagnostics beneath the star, not the star itself. The discipline is refusing to let a diagnostic get promoted to a goal just because it happens to be easy to watch.
Pick the one number that, if it moved in the right direction, would mean your customers' lives got easier. Put that on the wall. Then make tickets-closed a means to it, not the thing you worship.