Ask four leaders in the same company what is wrong with the technology, and you will get four different answers — each delivered with total conviction. Engineering says the architecture is fighting them. Product says the roadmap keeps slipping. Finance says the cost line is growing faster than the plan allows. Support says the defects never stop coming. Each is describing something real. Each is also describing the same thing.
This is the parable of the blind men and the elephant, played out in a leadership meeting. One holds the trunk, one the leg, one the ear, one the tail. Every description is accurate. Every description is incomplete. And because each function is rewarded for solving its own version, the company funds four separate cures for one disease — and is surprised when none of them work.
The reason this matters is not philosophical. It is the single most common way a CEO ends up paying for the wrong treatment: by mistaking the loudest symptom for the disease, and the most credible executive for the most correct one.
Four seats, four views, one underlying disease
Consider a company whose real disease is a tangle of tight coupling and missing service boundaries in the core platform — an architecture problem in the operating-model layer. Here is how that single disease presents in four different rooms:
- Engineering sees architecture friction. Every change ripples through three systems. Velocity is real, but most of it is consumed fighting the structure rather than shipping value. To engineering, the disease is technical debt.
- Product sees roadmap pressure. Commitments slip, scope shrinks, and the backlog grows faster than it clears. To product, the disease is delivery — and the obvious fix is more people or a tighter process.
- Finance sees rising cost. Headcount climbs, the cloud bill climbs, and output does not climb with them. To finance, the disease is efficiency, and the obvious fix is a hiring freeze or a cost programme.
- Support sees defects. The same categories of issue recur because every fix risks breaking something adjacent. To support, the disease is quality, and the obvious fix is more testing.
Four functions. Four diagnoses. Four budget requests. And not one of them names the actual disease, because no single seat can see the whole animal. Technical debt, slipping roadmap, rising cost, and recurring defects are not four problems. They are four shadows cast by one.
Why the loudest symptom usually wins — and usually loses
In the absence of an independent diagnosis, companies do not resolve these competing views through analysis. They resolve them through politics. The function with the most organisational power, the most articulate leader, or the most board attention gets its version funded. The disease is named by the loudest voice, not the clearest evidence.
This is how a coupling problem becomes a "we need to hire more engineers" problem, or a "we need a stricter product process" problem, or a "we need to cut cloud spend" problem — depending entirely on who was most persuasive that quarter. The treatment follows the org chart instead of the diagnosis.
When the disease is named by the loudest function rather than the clearest evidence, the company funds the cure with the best lobby, not the one that works.
And it fails predictably. You hire more engineers into a tightly coupled platform and they slow down too — the architecture absorbs them. You impose a stricter product process onto a structure that cannot absorb change cheaply, and you have added ceremony to friction. Each function's local fix is rational from its seat and useless against the disease, because the disease lives in the space between the seats.
Diagnosis is the act of connecting the views
Here is the definition that reframes the whole problem: diagnosis is not the discovery of a new fact that the functions missed. It is the act of connecting the partial, accurate views that each function already holds into one coherent picture of the underlying disease.
The engineer is not wrong about the friction. The CFO is not wrong about the cost. The product lead is not wrong about the slippage. Each is holding a true fragment. The diagnostic work is to lay those fragments on the same table, recognise that they are surfaces of one structure, and name that structure precisely. This is the core of how the three frameworks operate: a Growth Blocker MRI Dx™ exists specifically to reconcile the four stories into a single named disease, before the Growth Accelerator Rx™ writes a single treatment that resolves all four symptoms at once.
When the diagnosis is right, something striking happens in the leadership team. The arguments stop. Engineering, product, finance, and support stop competing for budget because they realise they were describing the same disease from different doors. The treatment is no longer a contest between functions. It is a shared plan.
Why it has to be independent
This is the part that is structurally impossible to do from inside. Every internal leader sees the disease through the incentives, history, and vocabulary of their own function. Not because they are biased people — because they are good at their jobs, and their jobs define what they can see. The CTO cannot un-know the architecture. The CFO cannot un-see the cost line. Each is, by design, holding one part of the elephant and trained to defend that part.
An independent diagnosis has no seat to defend. It is not optimising for engineering's credibility, product's roadmap, finance's targets, or support's metrics. Its only loyalty is to naming the disease correctly — which means it can say the thing no internal function is positioned to say: that all four of you are right, and all four of you are looking at the same problem.
That is the value of an outside view. Not more intelligence than the team has. More vantage than any single seat can have. The functions supply the fragments; the diagnosis supplies the picture. And the company stops paying for four cures to one disease.
When your functions disagree about what's wrong with technology, they are usually all correct and all incomplete — the job is not to pick a winner but to commission a diagnosis that connects their views into one named disease.