Where we are
Five sessions of building capability. This one starts the harder half of the track, and it is the session that determines whether the rest of it goes well. Here is the pattern to break: a manager notices output is below the bar, feels responsible, and reaches for whichever tool is nearest - usually a serious conversation, a warning, or a form. All three are interventions for a will problem, and will is only one of four causes. Applied to a skill gap, a warning is cruel and useless. Applied to a context problem, it is worse than useless - it holds someone responsible for a decision made above them. The examination costs you one conversation and about twenty minutes of thought, and it routinely changes the answer completely.
Four causes, four different answers 7 min live
Run the questions in this order. The order matters: skill and context are the two most often missed, and both get misread as will, which is the one cause managers reach for first.
LiveThe questions that actually separate them3 min▶
- Skill or will? "Could they do it if their life depended on it?" If yes, it is not skill. Also: have they ever done it well? A person who did this competently last year has the skill and something else changed.
- Context or person? "Is anyone else in a similar position struggling in the same way?" If two of three people on the same work are behind, it is the work. Also: what changed, and when? Line the timeline up against reorgs, system changes, manager changes and departures.
- Fit or skill? "Would six months of good teaching plausibly fix this?" If the honest answer is no because the job needs something they fundamentally do not have or want, it is fit - and fit is not a character flaw, it is a matching error, often one you made at hiring or promotion.
- The will question, asked properly: not "why are you not trying" but "something has changed - can you tell me what?" Disengagement almost always has a cause: a passed-over promotion, a manager change, a personal situation, burnout, or a reorg that removed the part of the job they liked.
The two-of-three test in action. A Cartwheel lead was preparing a serious conversation with an analyst whose reports were consistently late. Before it, she checked: two of the three analysts on that reporting cycle were also late. The cause was a data feed that had started arriving four hours later after a systems change, and none of them had mentioned it because each assumed it was their own problem to absorb. One question, asked before acting, prevented three unfair conversations and found an actual fix.
The wrong intervention is not neutral 5 min live
A wrong diagnosis does not merely fail to help. Each mismatch has a specific, predictable damage profile, and three of them actively make the situation worse than doing nothing.
LiveMixed causes, and where to start3 min▶
Real cases are frequently two causes stacked, and they stack in a predictable order - a context change creates a skill gap, the person struggles, nobody says anything, and six months later they are disengaged. By the time you look, you see the will problem, because it is on top.
- Peel from the bottom. Find the earliest cause on the timeline and address that first. Fixing the will problem while the context problem remains produces a brief improvement and then a relapse.
- Expect the will layer to resolve itself more often than you think once the underlying cause is fixed. Much apparent disengagement is a rational response to an unfixed obstacle.
- Name the sequence out loud with the person: "I think the role changed under you in March, nobody taught you the new part, and you have understandably lost heart. Let us start with the second thing." That sentence repairs a lot on its own, because it tells them you are not holding them responsible for the whole chain.
Self-studyRuling yourself out as the cause2 min read▶
Context includes you, and it is the check nobody runs on themselves. Four honest questions:
- Have they had my time? If 1:1s have been cancelled for two months, some of this is mine.
- Have I been clear? Would a third party reading my instructions know what good looked like? Or did I communicate the standard once, verbally, in a busy week?
- Am I giving contradictory priorities? Two urgent things and no ranking means whatever they choose will be wrong, which looks like poor judgment and is actually mine.
- Do they feel able to tell me something is not working? If not, I will always find out late, and I will always find out from the output rather than from them.
If two or more of these land, run the fix on yourself for six weeks before you conclude anything about the person. It is also the answer you would want a manager to reach about you.
Diagnose four Cartwheel cases ★ 14 min · everyone tries
Four real-shaped situations. For each, run the four questions in order, name the cause, and name the intervention. Do not skip ahead to the intervention - the discipline is the point.
Case A. Priya's forecasts are late and need rework. She is visibly trying, asks good questions, and has never done forecasting anywhere before.
Case B. Tomas delivered strongly for two years. Since the reorg he misses commitments without flagging them and has gone quiet in meetings.
Case C. Dana's role was rewritten around a new system, the team lost two people, and her output has halved. She is working longer hours than anyone.
Case D. Ade is excellent with suppliers and analysis, and was promoted to lead the team six months ago. Team output is flat, two people have complained about micromanagement, and he is doing the interesting work himself.
| Case | Cause | The tell | Intervention |
|---|---|---|---|
| A · Priya | Skill | Never taught it, and trying hard | Instruction with a named teacher, then review. This is b4's growth goal, not a warning |
| B · Tomas | Will, with a cause | Did it well for two years, so the skill is there. Something changed at the reorg | An honest conversation about what changed, before anything formal |
| C · Dana | Context | Working harder than anyone, and the job is a different job now | Fix the role or the resourcing. A plan here would be the b7 error |
| D · Ade | Fit - for the manager role | Excellent at the old job, mismatched to this one. Nobody taught him either | Both: teach the manager skills, and be open that the IC track may be the better role |
Examine one live case of your own ★ 8 min · do your own
Take a real situation you are currently worried about and run the examination properly before you do anything else with it.
Write the timeline first: when did output change, and what else happened around that date?
Run the four questions in order and write the answer to each, with evidence.
Run the two-of-three test: is anyone else in a similar position showing the same pattern?
Run the four questions about yourself.
Now name the cause and the intervention - and if the honest answer is "I do not know yet", the next step is one diagnostic conversation, not a plan.
"I do not know yet" is a legitimate diagnosis. The most useful outcome of this examination is often not a cause but the discovery that you do not have enough information - typically because you have been inferring everything from output and have not actually asked. The correct next step then is a single open conversation: "I have noticed the reports slipping over the last two months. I do not have a view on why yet, and I would rather hear it from you than guess." That sentence is available to any manager, costs nothing, and is the step most commonly skipped on the way to a formal process that need not have happened.
Try it yourself - this week ◐ 30-45 min total
- Run the full examination on one live case, including the timeline and the two-of-three test.
- Run the four self-check questions honestly. If two land, fix those first for six weeks.
- For anything you have been calling a motivation problem, check whether the person ever did it well. If they did, look for what changed.
- Have one "I do not have a view yet, I would rather hear it from you" conversation.
- Check whether any current concern started at the same time as a reorg, system change or departure. Line the dates up.
- If you conclude a case is fit, ask whether the mismatch was created at hiring or at promotion - and what that says about the process, not the person.
Sources covered (cited)
Taught from the management canon and improvement-plan practice. This session covers:
Three questions before you go 🎯 ◐ 90 seconds
1 · Why should "have they ever actually been taught this?" be the first question?
Ask "were they taught, by whom, when?" - not "should they know". A serious conversation, a warning and a form are all interventions for a will problem, and will is only one of four causes.
2 · Two of the three analysts on the same reporting cycle are late. What does that tell you?
The two-of-three test separates context from person in one question. In the Cartwheel case a data feed had started arriving four hours later after a systems change, and each analyst had assumed it was their own problem to absorb.
3 · A case looks like disengagement, but the timeline shows the role was rewritten six months earlier and nobody taught the new part. Where do you start?
Causes stack, and will sits on top because it arrives last. Peel from the bottom, and say the sequence out loud - it tells the person you are not holding them responsible for the whole chain.