Who this is for
This is a steering tool, not an execution tool. Its reader oversees several initiatives, meets on a weekly or monthly rhythm, and makes a small number of decisions that matter: what to protect, what to stop, where to ask a hard question. Each dot is a commitment the organisation has made, not a working to-do list.
So the record is deliberately short. A commitment carries evidence, a priority, clarifying questions that steering has asked of its owners, and blockers that need steering's help. Next steps, tasks and day-to-day decisions belong in the teams' own tools.
An execution view of the same work would use the hill for a different job: daily updates by owners, with a record that is expected to be incomplete. Here, completeness of evidence is the point.
Two questions status never answers
Every status report in a large company answers "how far along is it?". Two better questions go unanswered. First: what do we actually know — is this work past its unknowns, or does it only look busy? Second: does it matter right now — is this the work that deserves the scarce attention of the people who own it?
A hill chart answers the first question in principle. But if the dot is dragged by hand, it answers with the owner's confidence, not with evidence. And the hill says nothing about the second question: a nice-to-have and a regulatory deadline look identical halfway up the slope.
The cost is concrete. Important work that is not yet urgent — the design that should be settled before a migration, the question nobody owns — gets crowded out by whatever is loudest this week. It surfaces months later, as a crisis.
Drucker observed that executives are pushed by the flow of events. Unless they deliberately decide what to work on, the urgent will always crowd out the important.
Options
We looked at three ways to place a dot and three ways to show priority, then chose the combination that needs the least ongoing effort and is hardest to fool.
| Placement | How it works | Strength | Weakness |
|---|---|---|---|
| Declared stage | Owner picks "designing", "implementing"… | Fast, familiar | Stage names are elastic; "designing" can mean anything |
| Activity from tickets | Closed tickets move the dot | Zero effort | Measures motion, not certainty |
| Evidence checklist | Nine yes-or-no statements, each worth a fixed distance on the hill | Hard to fake; says what is missing | Needs agreement on the statements |
| Priority | How it works | Strength | Weakness |
|---|---|---|---|
| Rank everything | A single ordered backlog | Unambiguous | Collapses across streams; constant re-ranking |
| Colour by RAG | Red, amber, green | Familiar to leadership | Mixes progress, risk and priority into one colour |
| Importance × urgency | Importance is declared; urgency is derived from the due date | Only one judgement to make; the calendar supplies the other | Importance can be inflated — so it is limited to three levels |
Proposal
Place dots by evidence
Each commitment carries nine statements. Five describe the uphill phase and together are worth exactly half the hill; four describe the downhill phase. A dot's position is the sum of the statements that are true.
One rule keeps this honest: downhill evidence only counts once every uphill statement is true. You can tick "build has started" on a commitment whose riskiest assumption is untested — the chart will flag it as building before it's figured out and leave the dot where the evidence says.
Stage names are still shown, but they are inferred from position rather than declared. That keeps the familiar vocabulary without letting it drift.
The weights are deliberately uneven. Testing the riskiest assumption is worth more than naming an owner, because it removes more uncertainty.
Rearrange the same dots by priority
Importance is declared at one of three levels: core to the stream's outcome, supporting, or nice to have. Urgency is not declared at all; it comes from the due date, rising steeply in the last two weeks. The prototype can morph the hill into an importance-and-urgency matrix, with every dot flying to its quadrant, so the two views are visibly the same work.
Four practices from The Effective Executive
- First things first. Important work that is not yet urgent and still uphill is listed as work to protect — especially if it has not moved for two weeks.
- Posteriorities. Drucker argued that deciding what not to do is the harder and more useful decision. Work that is neither important nor urgent is offered for setting aside; set-aside work leaves the roll-up but stays visible.
- Concentration. An owner carrying three or more unresolved problems at once is flagged: the uphill phase needs sustained attention, and fragments of time don't resolve unknowns.
- Contribution. A commitment without an outcome statement is flagged with Drucker's question: what will this contribute?
Streams, weighted by what matters
A stream's position is the importance-weighted average of its commitments. It cannot pass the crest while any part is uphill: until then it shows how much of the stream is figured out.
When a person disagrees
Dragging a dot is allowed, and it is itself a steering act: it says I don't believe the evidence. The dot gains a dashed ring; hovering or tapping it shows where the evidence places it. If it lands far from the evidence, the prototype shows the evidence list and asks the honest question. Dragged back: which of these statements is no longer true, or what work is missing? Dragged forward: which statement is true but unticked? The correction goes into the checklist, and the dot and the evidence meet again.
Clarifying questions and blockers
Steering's main tool is the question. Each commitment keeps the clarifying questions steering has asked, with their answers once given; a question unanswered for a week is surfaced. Blockers are recorded where steering can help remove them, and draw a short red bar in front of the dot. A blocker older than two weeks is surfaced too.
A steering agenda, not a red status
Some commitments are likely to be broken unless steering acts: due within three weeks with the riskiest assumption untested, past their date, blocked for two weeks, building before they are figured out, or moved back over the crest by someone who doesn't believe the evidence. These go on a short agenda, at most five at a time, ranked by importance. The same trouble on nice-to-have work goes to decide what not to do instead: in trouble and not important is a reason to stop, not to escalate.
Every agenda item needs a response: re-date, re-scope, unblock, stop, or accept the risk with a reason. The response takes it off the agenda until something changes — the evidence moves, the date changes, the blocker ages another two weeks, or four weeks pass — and then it returns with its history. Items don't sit red to be defended in meetings; they ask for a decision, and every decision lands in a log that copies straight into the minutes.
Urgency is shown by a small flame between a dot and its name, so that rings mean only one thing: importance.
Seeing the work by system
Commitments are how steering thinks; systems are where they collide. Every commitment already lists the system components it touches, so a fourth view regroups the same dots by system: one thin hill per system, most at-risk first. Selecting a system highlights everything that touches it on the main hill and in the matrix, so a steering meeting can ask what is changing the order service right now? in one click.
One collision is serious enough for the agenda: two important commitments touch the same system, one is already being built and the other is still being designed. Whatever the second one decides may undo what the first is building. Steering can re-date or re-scope to separate them, accept the risk, or name a design authority who owns the system's design until both are past the crest. Three or more commitments still uphill on one system is a milder concern, listed as a crowded system. And ticking affected systems identified with no systems listed is flagged, to keep that statement honest.
History and sharing
A snapshot is saved automatically each day. Comparing against an earlier snapshot draws each dot's movement. The chart copies as an image and the status as a plain table, both of which paste cleanly into Confluence and email.
Limits
The nine statements are a starting vocabulary, not a standard; teams will want to tune the wording and weights. Importance remains a judgement and can be inflated. Data lives only in this browser; import and sharing are next steps.