Where we are
Every session handed you a component: the lifecycle and the four causes in b1, the written bar and the ramp in b2, GROW and maturity-matched oversight in b3, the skills matrix and the growth goal in b4, the translated ladder and the running evidence file in b5, the diagnostic sequence in b6, the six plan components in b7, the four beats and the two justices in b8, the exit floor in b9. Assembled badly, those parts still produce a manager who is busy and a team that is not developing. So this session is about the joins, and two of them carry most of the weight. Everything must point at the same standard - the bar in your ramp, the criteria in your ladder and the expectations in your coaching have to be the same thing, or people are being measured against a moving target. And every part must survive someone who wants to look good rather than be good, which is the pass most systems skip.
Seven components, and what each gap costs 7 min live
Read it as a sequence: a person joins, is taught, is coached, grows, is judged fairly, and - occasionally - leaves well. Each component has a job, and each absence has a symptom specific enough to diagnose from the outside.
LiveKeep it to one page, and make it point at one standard3 min▶
Two constraints that do more for quality than any additional process:
- One page, or it does not exist. A system your team cannot hold in their heads is a document you will reread annually. One written bar per role, one ramp, one cadence, one growth goal per person, one translated ladder, four diagnostic questions, one escalation route. If it will not fit, cut until it does.
- One standard, three places. The bar in your ramp, the criteria in your ladder and the expectations in your coaching must be the same words. When they differ, people are being measured against a target that moves depending on which conversation they are in - and that is experienced as unfairness even when every individual judgment was reasonable.
The test for the second one is quick: read your ramp document and your ladder level side by side. If a new joiner meeting every ramp milestone would still not meet the entry level of the ladder, one of the two is wrong.
Self-studyWhat you own and what you escalate2 min read▶
Some of this system is not yours to fix, and knowing the line prevents both learned helplessness and pointless heroics.
- Yours entirely: the written bar for your roles, the ramp, the coaching cadence, growth goals, the translated ladder, the evidence file, the diagnosis, and the quality of every conversation.
- Yours to escalate: a span of control that makes coaching arithmetically impossible, a ladder with no dual track, no internal redeployment route for fit cases, a plan template with no support component, and a promotion process with an undefined potential axis. All of these are the leader track's sessions, and all of them are legitimate things to raise with evidence from your own team.
- The trap: treating an escalation item as a personal failing and quietly absorbing it. A manager with fourteen reports who blames themselves for not coaching enough is misdiagnosing exactly as badly as they would be with a person.
Three hostile readings 5 min live
A checklist rewards good intentions. Red-teaming does not. Take each role in turn and genuinely try to break your own system - and remember that you are attacking the design, never naming a person who might do it.
LiveThe single question that tells you if it works3 min▶
All seven components can be in place and the system can still be dead. The test is not whether the process ran; it is whether the information reaches you early. One question:
Is it safe, on my team, to say "I am struggling with this"?
- If yes, the leverage curve from b1 works. Problems arrive in week three, when they cost a conversation. Diagnosis is easy because people tell you what is actually happening. Growth goals get chosen honestly rather than performatively.
- If no, every component degrades. Ramp problems hide until they are performance problems. Coaching becomes status reporting. Diagnosis runs on inference because nobody volunteers the cause. And the first honest signal you get is the output, months late.
- The evidence is behavioural, not survey-based. When did someone last tell you they had made a mistake before you found it? When did someone last say they did not understand something in a meeting? If you cannot remember, you have your answer.
Edmondson's point holds precisely here: this is not softness. High standards plus safety is where people improve. High standards plus fear produces concealment, and then formal processes.
Assemble the Cartwheel merchandising system ★ 14 min · everyone tries
One team, four people, all seven components on one page. Build it in order, and notice how much of it already exists in the sessions you have done - the assembly is mostly retrieval.
Write the bar for each role, in the language of the ladder level below it, so the two match.
Write the ramp once, reusable, with named teachers.
Set the cadence, with oversight matched per person-and-task rather than per person.
One growth goal each, riding on real work, from the skills matrix.
Name the diagnostic discipline and the escalation route, so that a concern has somewhere to go that is not a template.
| Component | Content | Rhythm |
|---|---|---|
| 1 · Written bar | Analyst and senior analyst, in the same words as the ladder | Reviewed yearly |
| 2 · Ramp | 30/60/90 observable milestones; teacher named per phase | Per joiner, checkpoint at 90 |
| 3 · Coaching cadence | Weekly 1:1, theirs not mine; oversight per person-and-task | Weekly, 30 min |
| 4 · Growth goals | Priya: forecasting. Tomas: pricing model. Dana: category planning. Ade: coaching skills | One per person, per quarter |
| 5 · Ladder + evidence | Translated once for merchandising; running dated file per person, shared with them | Four minutes a month |
| 6 · Diagnosis | Taught it? Conditions? Fit? Only then will. Plus the two-of-three test | Before any intervention |
| 7 · Escalation route | Concern in writing at the time; HR consistency check before any formal step | As needed, never skipped |
| Risk noted | Pricing model: only Ade can teach it. That is why it is Tomas's growth goal | Reviewed quarterly |
Red-team what you just built ★ 8 min · do your own
Three passes on a real system - ideally yours, with a peer running the first one. Fix each hole at the component that caused it, not at the symptom.
Optimiser pass: find the cheapest route to looking good. Write it down, then close it.
Person pass: find where someone could be judged for something outside their control, or be blindsided.
Auditor pass: ask for criteria, evidence, decision-maker and consistency check. Note everything that exists only verbally.
Fix at the component, then re-read the whole page and check it still fits on one.
| Role | Attack found | Component at fault | Fix |
|---|---|---|---|
| Optimiser | Look busy in the 1:1 by bringing problems and never decisions | 3 · cadence | Ask "what would you do if I were away?" as a standing question |
| Optimiser | Pick a growth goal that is already nearly done | 4 · growth goals | Goal comes from the matrix gap, agreed with the teacher, not self-nominated alone |
| Person | "My report is late because the data feed moved, and that will look like me" | 6 · diagnosis | The two-of-three test, run before any concern is raised |
| Person | "I could be judged on a ladder criterion I have never been given work for" | 5 · ladder + access | Evidence file shared, and the manager owns making the work available |
| Auditor | The bar for senior analyst lives in the manager's head | 1 · written bar | Write it, in the ladder's words. Fifteen minutes, once |
| Auditor | No record of when concerns were first raised | 7 · escalation | Every concern in writing at the time, same day, brief |
What the whole track is actually for. The real test of a people system is not the quarter everything goes well. It is the first time someone on your team is genuinely struggling. A system that works produces a useful early conversation, an honest diagnosis, teaching if that is what is needed, and - occasionally - a fair parting that surprises nobody. A system that does not produces silence for nine months, then a template, then an exit that the team reads as a warning about themselves. Same organisation, same policy, same person: the difference was decided months earlier by whether the written bar existed, whether the coaching happened, and whether it was safe to say "I am stuck". Build for that, and the hard sessions in this track mostly stay unused.
Build your own - this quarter ◐ 60-90 min total
- Write your seven components on one page. If it will not fit, cut until it does.
- Read your ramp bar and your ladder level side by side. If a joiner could meet every ramp milestone and still miss the ladder's entry level, fix one of them.
- Ask a peer to run the optimiser pass on your system. You cannot find your own shortcuts.
- Run the person pass and the auditor pass yourself, and write down everything that exists only in your head.
- Separate what you own from what you need to escalate, and raise the escalation items with evidence from your own team.
- Answer the last question honestly: when did someone last tell you they were struggling, before you found out from the work? If you cannot remember, start there.
- Read the leader track (a1-a6) if you want to see the same system from the side that funds and governs it.
Sources covered (cited)
The capstone draws on the whole course. This session covers:
Three questions before you go 🎯 ◐ 90 seconds
1 · Your strongest people keep leaving saying "there was no path here". Which component is missing?
Each gap has a diagnosable symptom. Components 1 to 4 are cheap and prevent most of the expensive problems in 6 and 7 - and almost everyone invests the other way round, tightening the exit process while starving the coaching layer.
2 · Why must the ramp bar, the ladder criteria and your coaching expectations use the same words?
One standard, three places. The quick test: if a joiner could meet every ramp milestone and still fall short of the ladder's entry level, one of the two documents is wrong.
3 · All seven components are in place. What single question tells you whether the system is actually alive?
Without safety, every component degrades: ramp problems hide until they are performance problems, coaching becomes status reporting, and diagnosis runs on inference. Check it behaviourally - when did someone last tell you about a mistake before you found it?