Depth of sight into work you have commissioned is a design parameter with a price. Most organisations never set it, and the few that buy it cannot always read what it shows them.
An airline outsources heavy maintenance to a third-party provider. The scenario is hypothetical; the arrangement is ordinary. Each month a report arrives: turn times, findings, rectifications, a schedule adherence figure, an amber flag on parts availability that has been amber for a while. Everything in the report is true. The airline's own people have not stood in the hangar since the contract was signed.
Ask an executive in that position what they can see and the honest answer is usually: whatever the provider chose to report. Ask when that was decided and there is rarely an answer — it was settled by the contract template and by a reporting cadence someone drafted during mobilisation.
That is the finding worth sitting with. How deeply a client can see into work it has commissioned is one of the earliest and most consequential decisions on any major engagement, and it is almost never taken as a decision. It is inherited.
The Strategic Context
Winch, Usmani and Edkins, in Construction Management and Economics in 1998, borrowed a term from the service-quality literature to name the parameter: the line of visibility — the level of transparency of the delivery process to the client. Their argument is that where the line is drawn should be an explicit early choice, made as seriously as the choice of contract form, because it determines what the client can learn and when.
Their case study makes the abstraction concrete. Studying a large pharmaceutical research campus in England — handed over, their paper reports, in March 1995 at a construction cost of around £550 million — they document a client that pushed the line unusually deep. It placed roughly a hundred of its own staff on the project through the construction phases, and those people took an active part in all the design reviews and the major site inspections. It also stood up an internal engineering quality control department of twelve engineers, experienced in the relevant process plant, to review the design consortium's output.
The authors report two results. The changes that department proposed were formally costed at a saving of between £4 million and £5 million, achieved without compromising the functionality of the design. And, more importantly for an executive reading this: because the client participated at all significant reviews, there was no way it could be surprised.
Being unable to be surprised is a strange-sounding objective. It is also, on the evidence of that research, a better description of what a client actually wants from governance than anything on a status report.
What Leaders Commonly Misread
That reporting and visibility are the same thing. A report is an account of the work produced by the party doing it, at a moment of their choosing, in a format designed to be defensible. Visibility is presence at the event where a decision is made. The first tells you what someone concluded; the second lets you watch the conclusion being reached, and to intervene while intervening is still cheap.
That more reporting closes the gap. It does not, and the reason is structural rather than cultural. Every layer that summarises discards something, and what gets discarded first is the anomaly — the detail that does not fit the pattern the report is organised around, which is exactly the detail a governing body needs. [Related article: Whose Knowledge Does Your Governance System Actually Hear?] develops that mechanism. The relevance here is narrower: a deep line of visibility is one of the few designs that bypasses the problem entirely, because the client is in the room rather than reading a summary of the room.
That visibility is free. It is not. The client organisation pays for it in its own people's time, and the research is candid that few clients — even experienced ones — have either the desire or the capability to push the line as deeply as its case study client did. Pretending otherwise produces the worst of both worlds: contractual rights of access that nobody exercises, which give the appearance of oversight and the substance of none.
That access is oversight. This is the misreading that gives the article its title. A client with the right to attend every design review and nobody able to interpret a drawing has bought observation. What it does with the observation is the whole question, and it is a capability question, not a contractual one.
Reframing the Issue
Read properly, visibility is a design parameter with two independent settings — and setting one without the other wastes the money.
The first setting is depth: which events you attend, which artefacts you see, which decisions you are present for. This is negotiable at contracting and expensive to change afterwards.
The second is capability: whether the people who attend can read what they are shown, and whether they carry enough standing to say so when it is wrong. This is a resourcing decision inside your own organisation, and it is the one that gets cut.
Set out that way, the choice is not between trust and oversight. It is between two coherent designs that buy different things.
Presence buys early knowledge and the ability to influence a decision before it hardens. It costs client headcount, and it carries a real hazard: a client inside the process can slow it, and can blur accountability for outcomes it helped shape.
Exception reporting buys cheapness and clean accountability. The delivery party is left to deliver and reports only when a bound is breached. It costs learning latency — you find out when the threshold trips, which is by definition after the fact. [Related article: Authority With an Expiry Date] is the fuller treatment of that design, including how bounds are set and what a breach obliges. The point here is only that the two designs are alternatives with different prices, and that most organisations end up with exception reporting by default while believing they chose oversight.
Strategic Analysis
Visibility earns its keep by finding definitional gaps, not by catching cheats
The intuition behind client oversight is usually adversarial: someone must watch the supplier. The evidence points somewhere more interesting.
The same case study records a genuine problem on that project — the parties could not agree what "complete design" meant. British and American practice differed; incompatible drafting systems meant issued drawings were not always coordinated; and performance specifications for services left real uncertainty about equipment sizes, so design in one package kept disturbing others. Nobody was at fault. The gap was definitional, and it would not have surfaced in any report, because each party was reporting accurately against its own understanding.
The resolution was itself a visibility move: up to seventy engineering staff from trade contractors and suppliers co-located with the design consortium to work it out. That is not surveillance. It is a line of sight deep enough to reveal that two organisations meant different things by the same phrase, early enough to matter.
Visibility's return comes overwhelmingly from this class of discovery — mismatched assumptions, incompatible definitions, decisions being taken on a basis the client would not have chosen — rather than from detecting bad faith. Executives who justify oversight on the second basis tend to under-invest, because bad faith is rare and they know it.
The capability half is the half that gets cut
The twelve engineers in that client's review department were experienced in the specific process plant being designed. That specificity is the point: not auditors, not contract managers, but people who could look at an output and know whether it was right.
Two findings in the same paper point the same way. A survey of eight leading UK property developers indicated a shift towards establishing in-house capability able to deal directly with designers and contractors. And separate research the authors cite found that project performance against time, cost and quality criteria was better where the client had an in-house architectural team.
Both are indicative rather than conclusive, and the paper treats them that way. But they line up with the mechanism: visibility converts into value only through people who can interpret what they see. An organisation that outsources the work and also outsources the ability to understand it has not transferred a risk; it has transferred its own judgement, which is a different transaction and a worse one. [Related article: Risk You Transfer Is Risk You Still Own] deals with what happens to the exposure itself; the question here is who is left able to look.
Building an internal review capability is a capability investment and should be appraised as one — [Related article: What Does the Enterprise Own After a Capability Investment?]. What it is for is narrower than duplicating the supplier's work: to be present, competent and early.
The conditions that made deep visibility work
Three features of that case deserve attention from anyone contemplating it, because deep visibility without them tends to produce friction rather than value.
The principal parties were engaged on a cost-reimbursement basis with open-book accounting, so the additional effort caused by client-driven changes was visible and recoverable. The client's own knowledge of the process was deep enough that when changes were needed, the discussion happened in what the authors describe as a trusting atmosphere rather than an adversarial one. And the client had already abandoned an earlier attempt at the same project, whose estimated outturn had moved from £500 million to £800 million — an expensive but effective way to acquire an appetite for looking closely.
The generalisable lesson is not the depth. It is that depth must be matched by a commercial structure that can absorb it. Deep client presence inside a fixed-price arrangement with no mechanism for compensating client-driven change produces claims, not savings.
Decision Framework
Make the line of visibility an explicit decision at approval, recorded in five lines.
| Line | The decision |
|---|---|
| Events | Which decision events will we attend — and which will we merely receive minutes from? |
| Artefacts | What will we see in its working form, rather than as a summary? |
| People | Who attends on our behalf, and what must they be competent to read? |
| Consequence | What may our attendee change, request or stop — and what happens if they raise a concern? |
| Commercial fit | Does the contract make our presence recoverable for the other party, or does it turn our involvement into a claim? |
Two tests.
The empty-chair test. For each event named, ask who would attend if that person were unavailable, and whether that substitute could read the same material. If the seat would go empty, or be filled by someone who could not interpret what was shown, the depth on paper is fictional — resource it or stop paying for it.
The surprise audit. After any unwelcome development on a major engagement, ask one question: at which event could we have learned this earlier, and were we in the room? A pattern of "we were not invited" is a depth problem. A pattern of "we were there and did not notice" is a capability problem. They have different remedies, and conflating them is why oversight budgets get spent on the wrong half.
Note the boundary with numeracy. Whether a person can tell that a claim put in front of them is arithmetically impossible is a distinct question, developed in [Related article: Enough Technical Depth to Test the Answer]. The two are complements: access without numeracy sees nothing, numeracy without access has nothing to look at.
From Strategy to Execution
Immediately. For your two largest outsourced or delivery-partner engagements, write down the current line of visibility as it actually operates — not as the contract permits. Most organisations find the two documents describe different arrangements.
Over two or three quarters. Resource the seats. A small number of genuinely expert people attending the events that matter beats a larger number of contract managers attending everything. Where the expertise does not exist internally, decide deliberately whether to build it, borrow it, or accept exception reporting and say so.
Longer term. Treat client-side capability as part of the sourcing decision rather than as an overhead discovered afterwards. The question at the point of outsourcing is not only can someone else do this better? but what must we retain in order to remain a competent client? An enterprise that never asks the second question will, over a decade, become structurally unable to evaluate the work it depends on.
Signals to Monitor
- Rights of access in the contract that have never been exercised. Depth on paper, exception reporting in practice.
- Client-driven change generating claims. Depth has outrun the commercial structure's ability to absorb it.
- Your own experts being asked to approve rather than to review. The seat has become ceremonial, and everyone in the room knows it.
Questions for the Leadership Team
- On our largest engagement, which decision events do we attend, and who decided that list?
- Of the people who attend on our behalf, how many could identify an error in what they are shown?
- When we were last surprised by a supplier, at which event could we have learned it earlier — and were we there?
- If we outsource this next capability, what must we retain in order to stay a competent buyer of it?
- Are we buying presence or exception reporting — and did we choose?
Closing Perspective
Every organisation that commissions work sets a line of visibility. The only variable is whether it sets it deliberately.
Set by default, the line lands wherever the standard contract and the standard reporting pack put it, which is shallow — and the client discovers what it needed to know at the moment the information has stopped being useful. Set deliberately, the line becomes a real choice between two honest designs: be present and pay for it in your own people, or stand back and accept that you will learn late.
What makes the choice consequential is the second setting. Depth without competence is the most expensive form of governance theatre available, because it costs contractual leverage, supplier goodwill and internal headcount, and returns nothing. The organisations that get value from looking are the ones that kept someone able to see. [Related article: Quality Assurance Cannot Tell You the Specification Was Wrong] covers what those events are entitled to decide; [Related article: Who Is Entitled to Say It Worked?] asks who is qualified to judge the result once everyone has gone home.
About EraNorth Insights
EraNorth Insights publishes practical analysis on strategy, projects, operations, transformation and decision intelligence for professional and organisational use. About EraNorth.