7 Signals That Suggest the Way You Run Is Breaking
Nothing has failed. Everything is a little harder than it was, and the explanations offered are always about the last thing rather than the pattern.
- Operating models rarely fail. They strain, and the strain shows as a set of symptoms that each get explained away individually.
- The seven signals below appear well before anything visible breaks. Each is easy to attribute to something else, which is why they usually run for years.
- The position worth holding is not “we are busy”. It is “we can tell the difference between growing pains and something that needs redesigning”.
Operating models do not usually fail. They strain, quietly, for a long time before anything visible breaks.
What a chief executive notices is a general sense that things are harder than they were. Decisions take longer. More meetings are needed to settle things that used to settle themselves. Good people are stretched. Nothing is wrong exactly, and everything requires more effort than it should.
Each individual symptom has a ready explanation. That project was unusually complex. That person was on leave. That quarter was difficult. Each explanation is true, and together they conceal a pattern.
The seven signals below are what that strain looks like from the top. None of them is proof of anything on its own. Three or more together usually means the way the business runs was designed for a smaller, simpler version of it.
More decisions are coming to you than used to
Things arrive on your desk that ought to have been settled two levels down. Not because your people are weak, but because it is genuinely unclear who decides, or because the decision crosses functions and nobody owns the crossing.
The natural response is to decide them, because they need deciding and you are quick at it. That works and it teaches the organisation to escalate, so the volume grows.
The signal is not the workload. It is that decision rights were set for a smaller business and never revisited. The fix is to write down who decides what, up to what value, and to defend that boundary even when deciding yourself would be faster.
The same issue gets solved repeatedly, in different meetings
A problem appears, gets discussed, gets resolved. Two months later a version of the same problem appears somewhere else and gets resolved again by different people who did not know about the first time.
Each resolution is competent. Nothing captures the cause, so nothing prevents the recurrence, and the organisation spends real capacity re-solving what it already solved.
This is one of the earliest signals and the easiest to miss, because every individual instance looks like ordinary problem-solving. The fix is to log recurring issues by cause rather than by incident, and to treat the top three as design work rather than as things to handle better.
Anything that crosses two functions takes a disproportionate amount of time
Work inside a single function moves reasonably. Work that spans two or three slows to a crawl, needs coordination meetings, and depends on personal relationships to get finished.
That is not a cooperation problem. It means nobody owns the process end to end, so it advances only when someone with enough goodwill pushes it.
The cost is invisible because it appears as diary time rather than as a line in the accounts. The fix is to name owners for the handful of processes that genuinely cross functions, with authority in all of them. Most firms have never done this, because each function is well run and the gaps between them belong to nobody.
Every symptom has a reasonable explanation. That is precisely why the pattern runs for years before anyone names it.
Your best people are the constraint on everything
A handful of people appear in every significant piece of work. They are excellent, they know how everything fits together, and nothing important happens without them.
That looks like strength and it is a structural risk. The organisation has concentrated knowledge in individuals rather than building it into how the work is done. Those people become the bottleneck on capacity and the single point of failure on continuity.
The signal is not that they are busy. It is that nobody could cover for them. The fix is deliberate: write down what only they know, spread the exposure, and design the work so competence rather than heroics is enough.
Three or more of these together usually means something needs redesigning rather than managing.
A free Executive AI Opportunity Assessment. Forty-five minutes, online, with your leadership team. We arrive with a model of how firms in your sector make money and where AI has moved those numbers — and you tell us where it is wrong.
Book a free assessmentThe numbers take longer to produce and mean less when they arrive
The management pack still gets produced on time. What has changed is the effort behind it — more manual assembly, more reconciliation between systems that disagree, more caveats explaining why a figure differs from last month.
And when it arrives it answers fewer of the questions being asked. The board wants a view by segment, or by channel, or over time, and the data cannot easily produce it.
The signal is the gap between what is asked and what can be answered without a week of work. It widens gradually, so nobody notices the trend until someone asks something ordinary and the answer takes days.
Well before anything visibly fails. Each one has a reasonable explanation at the time.
A hard quarter. A complex project. Someone on leave. Each explanation is true individually.
Good people work harder and the symptoms recede, which removes the pressure to redesign.
A resignation, a lost customer, a failed programme. The pattern becomes obvious in hindsight.
This describes a pattern rather than a measured dataset. The test worth running is your own: ask your leadership team to mark which of the seven they recognise, privately, before discussing it. The spread of answers is usually more revealing than the total.
Good people are leaving for reasons that sound reasonable
Departures happen. Each one has an explanation that is easy to accept — a better offer, a career move, a personal reason, a wish for something different.
What is worth looking at is whether the explanations cluster. People leaving from the same function. People leaving at the same tenure. People leaving after a particular kind of frustration. Exit conversations that all mention effort, or ambiguity, or not being able to get things done.
The signal is not the rate. It is the pattern in the reasons, which is only visible if someone looks across them rather than at each in turn. Most firms handle exits individually and never assemble the picture.
Three things worth doing this quarter
- Track for a fortnight which decisions reach you and which of those should have been settled below. The list is usually longer than expected and the pattern in it is the finding.
- Ask your leadership team to mark the seven signals privately, then compare. Where they disagree is where the strain is least visible from the top.
- Look at the last six departures together rather than individually. If the reasons cluster, that is a structural signal rather than six unrelated events.
Every explanation is about the last thing rather than the pattern
The most structural of the seven, and the one that hides the other six. When something goes wrong, the account offered is specific and local. That project was unusual. That client was difficult. That month was exceptional.
Each account is accurate. What is missing is anyone asking whether these things keep happening for a reason. That question requires standing back, and nobody in the operational rhythm has time to stand back.
So the pattern goes unnamed until an event forces it — a resignation that cannot be covered, a programme that fails, a customer lost for reasons that turn out to be internal. The fix is a deliberate look at the pattern rather than the incidents, scheduled rather than triggered.
What this means
None of these seven mean anything on their own. Every one has a reasonable local explanation, and most of the time the local explanation is correct. That is exactly why the pattern survives — it is made of individually defensible events.
What makes them worth watching together is that they share a cause. The way the business runs was designed for a smaller, simpler version of it. Decision rights, process ownership, knowledge, reporting and capacity were all sized for the firm as it was. The business grew and none of them were revisited, because nothing forced the question.
The economics of acting early are strongly favourable. Redesigning deliberately, over a year, costs planning time and some organisational discomfort. Doing it after an event forces the issue costs the event as well — and events of this kind tend to be a lost key person, a failed programme, or a customer who left for reasons nobody attributed correctly.
Which is the same argument we make about technology. Change the work before you change the technology — and if you are about to put AI on top of how the business currently runs, be certain about how it currently runs, because AI will speed up whatever it finds.
Growing pains, or something that needs redesigning?
Forty-five minutes with your leadership team, online, at no cost. We bring a working model of how firms in your sector make money and where AI has moved those numbers. You tell us where it is wrong. Then we tell you what we think — including “don’t invest”, if that is the honest answer.
Book a free assessment45 minutes · online · no cost · nothing to prepare
Continue reading
9 Signs Your Firm Has Outgrown the Way It Is Governed
The same growth, showing in the formal scaffolding rather than in daily operations.
For chief executives9 Hidden Operating Frictions Reducing Customer Profitability
What the same strain costs, measured at the customer.
For the whole board9 Reasons Transformation Programmes Fail After Approval
What happens when you try to fix it with a programme instead.