Work people grow in

When someone sounds flat about their work, a lead's first guess is often motivation: maybe they've lost interest. Often the real cause is the way the work is cut.

Job design means shaping what a role contains, so people can do it well and grow in it. It's a lever a lead controls directly, without a budget or a reorganization.

At Klasna, Maryna says at her one-to-one: “I answer the same three questions forty times a day. I don't even know if the families' lessons went well after.” That's not a motivation problem to fix with a pep talk or a bonus. It's a design problem: her work has been cut into small identical fragments, with no end she ever sees and nothing that uses her other skills.

The fix isn't to remove the routine tickets: in a peak they have to be answered. It's to change how the work around them is cut. The next cards show what to change.

Research links five features of work to motivation and satisfaction:

1. Variety of skills: the work uses more than one ability.
2. A whole piece of work: you do something from start to finish, not a fragment.
3. Work that matters to someone: you can see who it helps.
4. Autonomy: you have a say in how and when you do it.
5. Feedback from the work itself: the work shows you how it went.

These are useful as a diagnostic, not a score. Look at a role and ask which features are missing. At Klasna, Maryna's March week has almost none: the same three questions (no variety), answered by whoever is on shift (no whole piece), with no sign of the exam behind each question (no visible meaning), and no idea how things turned out (no feedback).

Two things are often mistaken for these features. A more senior title changes the label, not the work. Constant supervision is the opposite of autonomy. Neither makes the day better.

In a peak you can't remove the routine tickets, but you can add the missing features with changes that cost no budget.

• A whole piece of work: when a family's problem needs three replies, one person follows it to the end instead of whoever is on shift.
• Work that matters to someone: the ticket shows “exam in 12 days”, not just “package question”. The same reply now has a reason.
• Feedback from the work: each person sees their own cases where the family wrote back, and their own reopens by cause.
• Variety: Olha's idea at Klasna, one hour a day off the queue for something else, such as the newcomer guide, a playbook or the AI pilot. Even an hour changes the day.

Each change is small, and each targets one feature. Together they turn “forty identical answers” into a day with an owner, a purpose, an outcome and a change of pace.

Start with the feature the person misses most; you'll learn which it is by asking.

Feedback from the work itself means people can see how their work turned out without waiting for a manager's opinion. It's one of the strongest of the five features, and one of the easiest to get wrong.

The right version at Klasna: once a week, each person sees their own resolved cases where the family wrote back, and their own reopens by cause. Their own numbers, in their own one-to-one. That tells Maryna whether the exam families got what they needed, and tells Taras which shortcut came back, in a way he can act on.

The wrong version is the league table: everyone's numbers posted side by side in the team chat. It turns feedback into a ranking, rewards volume over resolution, invites people to game the metric, and makes the quiet and the new feel exposed. It's the same trap as the ranking line in the Sunday-night plan from Case: a 30-60-90 plan for the Support team.

Feedback from the work should feel like a mirror, not a scoreboard.

Once a lead sees the five features, the temptation is to enrich every job at once: a new project for everyone. Ask first. Not everyone wants more scope, and more responsibility isn't automatically motivating.

At Klasna the idea comes quickly: Bohdan could lead the AI pilot! But at his one-to-one Bohdan said what he wants is fewer interruptions: two calm hours a day for the video-failure tickets. For him, good design is protecting depth, not adding scope. A project nobody asked for, added to a peak week, is extra load, not growth.

For Maryna the answer may be different: the refund work from Levels of freedom: matching person and task is something she's good at and wants to grow into.

That's the point of asking. People differ in what they want from work right now, and the same change can energise one person and exhaust another. The one-to-one is where you find out, before you design anything.

A stretch assignment is a task slightly beyond what someone has done, given on purpose, with support and a safe way to fail. Added on top of a full week, the same task becomes overload, and overload teaches the opposite of growth.

The rule: when something goes in, something comes out.

At Klasna, Maryna takes refunds as her stretch in the peak. What makes it growth: fewer schedule-change tickets for her (the AI drafts will speed those up anyway), her hour off the queue goes to refunds, and Olha approves her proposals for a month, which is the safe way to fail.

Three versions turn it into something else. Refunds in the evenings, after her queue: that's overload. Level 3 straight away to show trust: that removes the support. A bonus for every refund she closes: that rewards closing over getting it right, exactly what a refund shouldn't reward.

Room, time and support are what separate a stretch from simply more work.

Even a well-designed stretch can tip into overload in a peak, and people rarely say so first. A lead has to watch for the signs and ask.

Typical signs: replies sent at midnight; “I'll finish it at the weekend”; a person who used to raise ideas going quiet; small mistakes from someone who rarely makes them. None is proof on its own, but each is a reason to ask about load in the next one-to-one, with a neutral question: “How does your week feel at the moment?”

If it sounds like overload, reduce the load: take something off the plate, move a deadline, protect the hour. A lead controls those levers. If it sounds like exhaustion or a health problem, reduce the load and point to a professional for the health side; you're the lead, not a doctor or a therapist.

What doesn't help: praising the midnight replies as dedication. That teaches the whole team that overload is what good looks like.

A stretch assignment works when it's offered in the right order:

1. Ask whether the person wants more, and of what. Without their interest, a stretch is just more work.
2. Define the stretch: what's new, the support, and a safe way to fail.
3. Make room: take something off their plate.
4. Check in early and often while it's new and fragile.
5. Agree the next step together once it's working: more freedom, the next stretch, or staying put.

The same logic, smaller, fits newcomers. At Klasna, Dmytro and Sofiia are here for four months and mostly on routine tickets, since that's their level for now. Give each a whole small piece too: following one family's case from first message to resolution, with a named person to ask. Seasonal people who learn something real tend to work better and come back next spring; Iryna notes that would save two hiring rounds.

How it fits together: diagnose the work with five features, add them with small changes, ask before adding, and keep stretch from becoming overload. The next lesson is the month's case: splitting the whole spring peak between six people and an AI draft assistant.