![Silence is the least reliable management information in any business, and it is the information...]()
Silence is the least reliable management information in any business, and it is the information owners act on most often.
Three entirely different conditions produce an identical quiet. The business is genuinely handling it. Somebody is absorbing a problem at a cost that appears on no report. The third is that somebody has decided not to tell, having correctly predicted how it would land.
The corrective action for each is incompatible with the corrective action for the others. Designed quiet needs leaving alone, and firms that respond to it with new reporting requirements convert a working system into an administrative one. Absorbed quiet needs a capacity question rather than a conduct question. Withheld quiet needs a leader to examine their own last three reactions to bad news, which is the least popular of the three and the only one that works.
The diagnostic is a single question and the phrasing carries all of it. Asking whether everything is fine returns the answer the question invites. Asking what has been handled in the past fortnight that did not need to come up returns a list, and the list is the finding.
Which of the three do you suspect you have?
Check my bio to learn more on what I do![Handing over a decision transfers the work the moment somebody picks it up. The information takes...]()
Handing over a decision transfers the work the moment somebody picks it up. The information takes considerably longer, and in many firms it never arrives at all.
The reason is that the routes information used to travel were informal and largely invisible. A conversation in a corridor. A copy on an email chain. Being in the room when a client rang. None of it was designed, all of it felt like knowing the business, and every one of those routes closes the moment the owner steps out of the room.
What replaces them is nothing, unless somebody builds it deliberately. The week that follows a successful handover is often described by owners as unsettling rather than satisfying, and the unease has no object, which makes it difficult to act on.
The instinct is to ask for a report. The better move is narrower and considerably harder, which is to decide in advance what the business should surface without being asked, and to keep that list short enough to read over a coffee.
What would you want to know within twenty-four hours, even on a week you were nowhere near the building?
Feel free to have a look at my bio on my profile to learn a little more about me. 😊![There is a cost carried by capable people operating without a stated boundary, and it appears on no...]()
There is a cost carried by capable people operating without a stated boundary, and it appears on no workload review.
The description is unusually consistent across roles and industries. Rehearsing justifications for decisions nobody has questioned yet. Drafting the explanation before the outcome is known. Running a low-grade background process alongside the actual work, which continues at a rate the person has stopped noticing.
What removes it is not encouragement and not additional confidence. It is a boundary with two halves.
The first half is a stated limit, ideally a number. Sign anything up to this figure without asking.
The second half is cover. If a decision made inside that limit turns out badly, the position stated publicly is that the decision was correctly made. Most leaders give the first half and forget the second, and the second costs nothing while doing most of the work.
People who receive both halves describe the same fortnight afterwards. Same clients, same volume, same problems, and a noticeably better week, because a piece of attention that had been allocated to self-defence came back.
A team that decides freely is not a team that stopped caring about outcomes. It is a team that stopped budgeting for its own defence.
Check my bio to learn more!![A firm owner logged twenty-three decisions across a fortnight in July, expecting a modest result....]()
A firm owner logged twenty-three decisions across a fortnight in July, expecting a modest result. Fourteen failed the information test.
The distribution is where the case turns. Three were precedent, being client onboarding sign-offs that had come to him since the firm had six people and still came to him at thirty-one. Four were permission, being two managers uncertain of their authority on fee adjustments after more than a year in the role, neither of whom had ever been given a number.
Seven were protection, and all seven came from the same team.
Every one traced to an afternoon in March 2024, when a manager repriced a job without checking and was corrected in front of other people. Nothing formal followed. The owner had considered the matter closed within a week. Twenty-nine months later it was producing seven escalations a fortnight, and everyone in that team could have described the incident in detail. Nobody had asked them.
The repair was a conversation rather than a document. The event was named in a team meeting, the public correction was owned as the owner's error, and the manager received a written pricing band that week.
Escalations from that team fell to two over six weeks, which is the correct number, since those two were genuine exceptions.
Check my bio to learn more. 👆![A routing rule that lives in a founder's head is not a rule. It is a preference the team has to...]()
A routing rule that lives in a founder's head is not a rule. It is a preference the team has to infer from outcomes, and inference is why the same four decisions keep arriving.
The instrument is deliberately small. One page, three columns, ten to twelve rows covering recurring decision types rather than individual decisions. Anything larger becomes a policy document, which gets filed and opened by nobody.
The columns are decide, consult and inform. One name in the decide column, never a committee, since a decision with two owners has none. The consult column stays empty far more often than firms expect, as consultation carries a real cost and a rule requiring four conversations on a routine matter has rebuilt the bottleneck with more people in it.
The inform column is where founder anxiety belongs. Wanting visibility of a decision is not the same as needing to make it, and conflating those two is the most common reason decision rights fail to move. Visibility costs a line in a weekly summary. Authority is expensive and should be given deliberately.
Two fields decide whether the page survives a month. A limit expressed as a number wherever a number applies, and an exception route agreed before the first awkward case arrives.
Three rows this week, not ten.
Check my bio to learn more. ☝️![A decision log without a sorting method is a list of complaints. Sorting one takes about fifteen...]()
A decision log without a sorting method is a list of complaints. Sorting one takes about fifteen minutes for twenty items and runs on four questions.
Three of the four measure volume. They identify how much of a founder's decision load was never theirs, and the first question alone usually accounts for more than half of any list.
The fourth question is the one that changes anything, since it identifies cause rather than quantity. Precedent, permission and protection are indistinguishable on a page and require completely different repairs.
Precedent is a routing habit nobody chose. The first instance set a pattern and the pattern survived every version of the business since. A written rule fixes it cheaply.
Permission is a capable person who has never been told where their authority stops. An unbounded permission works exactly like no permission, since ambiguity at the boundary sends everything upward.
Protection is different in kind. The person could decide and is allowed to decide, and escalates anyway, because deciding independently went badly for someone at some point. No rule reaches that. The event has to be named.
Firms that read a whole log as a process problem write an elegant rule and watch a third of the volume continue unchanged.
Check my bio to learn more!![Clearing a decision backlog produces a genuine result and a second problem, usually about a...]()
Clearing a decision backlog produces a genuine result and a second problem, usually about a fortnight later, when the page has filled up again.
A business generates decisions continuously and routes them by habit rather than by design. Working through a backlog changes the contents of the queue. It changes nothing about where new items go, which means a founder who has successfully moved delivery work off their desk can find they have simply become the decision layer instead. Same occupancy, better disguise.
The measurement that opens this up takes ten working days and requires no change in behaviour at all. Every decision that arrives goes on a list. What it was, who brought it, roughly how long it took to settle. Nothing gets altered during the count, since changing conduct while measuring conduct produces a flattering number rather than a usable one.
One firm owner running this in July finished with nineteen items across ten days. He had estimated thirty-five, which is the usual direction of the error. Owners overestimate the volume and rarely know what the volume contains.
The count is not a measure of how much a founder carries. It measures where a business learned to send things, and it learned that early.
Want to read more? Check my bio.![The quality that makes a founder good at building a business is the same quality that keeps the...]()
The quality that makes a founder good at building a business is the same quality that keeps the decisions at the bottom of their list there for three years.
Being the person who can hold everything at once is a real capability, and it is usually why the firm exists at all. It also means the business learned very early that decisions route to one person. It kept routing them. That person kept absorbing them, and none of it ever produced a moment with enough room in it to make the two decisions that actually mattered.
The reframe worth carrying into next week is that a backlog is not evidence of avoidance. It is evidence of capacity, all of it spent, for years, on being the point everything passes through.
Strategic value is not a quality a leader develops. It is a set of decisions only they can make, sitting in a queue, waiting on a condition that was never about capability. The condition is room.
Which means the strategy has probably already been written. There has simply never been an afternoon in which to read it.
Start at the bottom of the list.![There is a cost to an unmade decision that appears on no workload review and in nothing anyone says...]()
There is a cost to an unmade decision that appears on no workload review and in nothing anyone says out loud.
People waiting on a decision about their own role describe the same experience with unusual consistency. Rehearsing conversations that have not happened. Reading meaning into scheduling. Recalibrating after every reasonable delay, and every delay is reasonable on its own terms.
The consistency extends to what happens afterwards. When the decision finally lands, the relief is rarely about the answer. It is about the absence of something that had been running quietly in the background for months. Same clients, same hours, same problems, and a noticeably better fortnight, because a piece of attention that had been allocated to a question came back.
Which reframes the register that this week has been about. An owner holding an unmade decision about a person is not holding a neutral item on a list. The cost is being paid weekly by someone else, in a currency that shows up nowhere.
The date a decision gets made is worth promising even when the answer is not. People manage themselves well on information, including disappointing information.
What are you carrying that belongs to somebody else's week?
More details are in my bio.![A firm owner protected one afternoon a fortnight from June and built a fourteen-item decision...]()
A firm owner protected one afternoon a fortnight from June and built a fourteen-item decision register. The distribution is ordinary and worth stating, since it repeats across almost every firm this exercise runs in.
Eight items were deferred purely for want of room. They took a combined ninety minutes and produced a real but modest result: a measurable drop in the interruptions reaching his desk, since each unmade decision had been generating its own small stream of questions.
Three were waiting on an input nobody owned. Each converted into an assignment with a name and a date in under five minutes.
Three had first arisen in November 2023. Whether the compliance-only service continued. Whether a senior operating at manager level for two years received the title. Whether a client billed at 2021 rates for roughly double the original scope was repriced or released.
Twenty-one months had passed. In every one of them, the service was sold, the senior operated without authority, and the client was billed at the old rate. Those were not pending decisions. They were decisions in full effect, described internally as under consideration.
Eleven items were worth clearing. Three were worth the afternoons.
Check my bio!![A decision backlog is three different things wearing one label, and treating the list as one thing...]()
A decision backlog is three different things wearing one label, and treating the list as one thing wastes most of the time recovered to address it.
Some decisions are deferred for want of room. No complexity, no cost, simply no uninterrupted half hour in which to make them. The delay cost is low per item and meaningful in aggregate, since each one generates small recurring friction across a team while it waits. The cause sits in Process, and the remedy is a batch rather than a session.
Some are waiting on an input nobody owns. A number nobody has calculated, a clause nobody has read, an intention nobody has asked about. These are not difficult decisions. They are decisions with an unassigned prerequisite, and they resolve the moment somebody is handed the prerequisite as a task with a date. The cause sits in Structure.
Some are deferred because making them costs something. These are the two or three items that have been on the list longest, and nothing external has ever blocked them. The cause sits in People, and usually in one person.
Sorting takes ten minutes. Not sorting costs about six months.![Protecting an hour is the part everyone plans for. What arrives in it is the part nobody describes...]()
Protecting an hour is the part everyone plans for. What arrives in it is the part nobody describes accurately.
The expectation is strategy: a clean sheet, a long view, three years out. What almost every owner actually meets in a first protected session is a list, and the list was already there.
A business generates decisions continuously and routes the ones only the owner can make towards the owner. With no room available, they do not disappear. They queue. Small ones get made quickly and badly in corridors. Larger ones get deferred, and a deferred decision still takes effect. It takes effect as whatever the business was already doing, remade every month, without anyone noticing a decision is being made at all.
One firm owner wrote his list in twelve minutes this June and produced eleven items. He had believed he was carrying three or four. Nine took under twenty minutes each. The two he did not touch had been sitting there since 2023.
That queue is the return on a handover, stated honestly. Hours are the mechanism. The decisions those hours finally allow are the result.
What would be at the bottom of your list?![The hardest question in this series is not how to hand something over. It is what a leader is for...]()
The hardest question in this series is not how to hand something over. It is what a leader is for once the handing over works.
An owner who is no longer operationally necessary has to answer that without the cover that being busy provided, and delivery work offers an answer immediately. It feels like contribution because it produces visible output. It costs the business the one seat nobody else can occupy.
The honest answer takes longer to sit with. A leader who is not required in the daily running has become the only person available to decide what the firm should stop selling, who needs to be capable of what by March, and what the next twelve months are genuinely being run against. None of that produces anything this week. All of it decides what the business is worth in three years.
The week's briefing tracks eighteen recovered hours to their actual destination, and the distribution is worth sitting with before the next transfer.
What would fill a full day your business gave back tomorrow, without checking a calendar and without saying "strategy"?![Two things happened in Australian business over the past fortnight that look unrelated and describe...]()
Two things happened in Australian business over the past fortnight that look unrelated and describe the same failure.
On 15 July the Prime Minister ended more than five years of voluntary AI governance, establishing an Office of AI and committing to a mandatory national framework, with National Cabinet considering the standards this month. On 30 July, research reported that Australian organisations had more than doubled AI spend year on year while fewer than a quarter had established testing or auditing processes for what the tools produce.
Capability was bought faster than the discipline to direct it was built. Regulation is now arriving to supply externally what the internal architecture did not.
The same error runs at firm scale every time a handover returns hours that nobody has claimed. One accounting firm logged eighteen recovered hours across six weeks: eleven to email, four to reviewing files already signed off by the person who now owned them, three to an overdue supplier contract, and none at all to the capability work that had been the stated reason for the transfer.
Eighteen hours delivered exactly as designed, strategic yield of zero. The transfer was not the failure. The absence of a claim on what it produced was.![The fear that follows a successful handover is rarely said out loud, and it sounds like this: if...]()
The fear that follows a successful handover is rarely said out loud, and it sounds like this: if the work runs without me, what exactly am I for?
Owners who name that early tend to design past it. Owners who leave it unnamed answer it by taking something back, usually within a month, usually something that had already transferred cleanly.
There are three destinations available to a recovered hour and only one of them is worth having. Reabsorption sends the hour back into the work just moved, and every individual instance of it is defensible, which is why it is nearly invisible while it happens.
Adjacency sends it sideways into other operational work, which feels like progress and produces none. Forward work is the third, and its signature is specific: it produces nothing measurable this week, it cannot be delegated without changing what the business is, and skipping it costs nothing today.
That combination is exactly why forward work loses every hour it has to compete for. It needs a claim of its own, made in advance.
Which of the three took the last hour a handover gave back?
