A delegation schedule states what a person may approve and says nothing about what that person may cause the enterprise to owe, because expenditure and obligation are fixed by two different systems and the enterprise writes only one of them.
Consider a hypothetical operations manager at a commercial diving and marine services firm, approval limit seventy-five thousand dollars. A client needs a subsea structure inspected inside a four-day weather window, so the manager writes: hold the window, we will have the vessel and crew alongside on Monday, paperwork to follow. Nothing has been approved; no limit has been exceeded.
Read as a scheduling courtesy this is trivial. Read as an outside party would — a record of what one enterprise told another and what the other then did — it is not. The client releases an alternative provider; the firm pulls a vessel off a second job; the inspection is then cancelled. The question is what the firm owes, and to whom.
Enterprises police this territory with one instrument, denominated in dollars of expenditure: the delegation schedule at the back of the governance manual. The exposure it is meant to contain is denominated in obligation, the two units do not convert, and no instrument in general use converts them. Does your delegation schedule govern obligation, or only expenditure? Most boards have never seen the difference, because nothing puts the two quantities side by side.
The Strategic Context
Delegation schedules are built from the inside out, calibrated against the enterprise's own balance sheet and appetite. They stop unauthorised spending and give internal audit a testable standard — both internal functions, both assuming exposure follows what the enterprise decides to pay.
Whether the enterprise owes anything is settled elsewhere: by a counterparty, on evidence about what its people said and did, in a forum it does not convene, under rules it did not write, after the fact.
An obligation is a finding another party goes looking for, and the material it searches is the enterprise's ordinary correspondence.
What the Schedule Was Never Built to Do
The first misreading is that signature is the moment of exposure. Signature is when the enterprise's own record-keeping engages; the counterparty's case is assembled from the weeks before anyone signed — the mobilisation instruction, the assurance about volumes, the note recording agreement on price.
The second is that preliminary instruments are inert. Letters of intent, heads of agreement and the rest of the pre-contract drawer carry a label chosen by the party that wants them non-binding; its weight is assessed against what both parties then did. An enterprise that behaves as though bound while insisting it is not has manufactured evidence, not protection. What such an instrument costs the other side — the free option it hands the buyer, priced into every later bid — belongs to [Related article: The Free Option You Give Away Before Anyone Bids]; this article owns only the gap between authority to spend and authority to bind.
The third is that exposure tracks seniority. Conduct does not respect that gradient: a scheduler who tells a supplier to start generates the same category of evidence as a director, and far more of it, because schedulers speak to counterparties daily and directors quarterly.
An approval right granted with no duty to examine what is approved is a distinct failure, carried by Article 54 in this collection. The concern here is prior: the right was sized against the wrong quantity.
Reframing the Issue
Stop asking what people may approve. Ask what they may cause the enterprise to owe. The schedule answers the first to the dollar and the second not at all, and only the second has ever ended an enterprise.
Two boundaries follow. An obligation that crystallises enters the reporting system on one set of rules and the bank account on another, and which ends an enterprise is the subject of [Related article: Accrual Governs the Report; Cash Governs the Outcome]; the question here is prior — whether anyone knew the obligation could be created. Once a contract exists, the live question becomes how much of the consequent failure the enterprise has agreed to absorb, a ceiling nobody aggregates across a supplier base and the territory of [Related article: The Liability Cap Nobody Adds Up].
There is also a level problem. Delegation is set enterprise-wide, the conduct that creates obligation happens on projects, and the only place the exposure could be added up is the portfolio — where nobody does, because that view is assembled from forecast cost and benefit, and an obligation created by conduct has neither until it arrives.
The Authority You Grant and the Authority You Cannot Withhold
The grant is bounded; the generation of exposure is not
An enterprise can define precisely what it confers on a role; the mechanics of a bounded grant are Article 21's subject in this collection. What it cannot bound is what its people generate: authority is given deliberately, in writing, in fixed amounts; evidence is produced continuously, informally, by everyone.
The evidence is a by-product of competent work
Mobilisation instructions, assurances about future volume, the silence after a supplier writes down what it believes has been agreed — these are artefacts of ordinary commercial diligence, and an enterprise cannot stop producing them without ceasing to trade. The control question is not how to prevent them but who may produce them, against which counterparties, and under what standing instruction.
The label on the instrument is not the control
Take a hypothetical festival promoter. A production coordinator with a ten-thousand-dollar limit writes to a lighting and rigging contractor: you are confirmed for the main stage, contract to follow. The contractor turns away two other events and books crew. Nothing was approved above limit; nothing was signed. The exposure is the contractor's displaced season, created by a sentence that read like good manners.
Decision Framework: The Obligation-Authority Map
The obligation-authority map plots every role holding a financial delegation against two coordinates, not one.
Coordinate one — the approval ceiling. Already in the schedule. Copy it across.
Coordinate two — the exposure ceiling. The largest obligation that role could cause the enterprise to owe in a single working day while staying inside its approval limit, derived from four inputs, each answered with a named counterparty and transaction, not a category. Mobilisation: the largest cost a supplier would incur if this role told it to begin. Displacement: the largest opportunity a counterparty would forgo on this role's assurance. Instrument: the largest value named in any document this role signs under a preliminary label. Silence: the largest assumption a counterparty has put to this role in writing and not been corrected on. Coordinate two is the largest of the four.
The threshold. Where coordinate two exceeds coordinate one, the role is uncontrolled in the only sense that matters, and a role with no approval limit at all can still carry a large coordinate two. A factor of one — exposure no greater than approval — is the only starting point a board can defend. Every role above the factor requires one of three changes, recorded with a name and a date: raise the approval ceiling to match what the role can already do; reduce the exposure through co-signature on mobilisation, standard-form correspondence, or a named list of roles permitted to issue preliminary instruments; or withdraw that role's unaccompanied access to counterparties.
The governance test. No delegation schedule is approved unless each expenditure limit carries an exposure ceiling beside it in the same currency, agreed by a named executive. A schedule with one column has answered half its question and reported the half as the whole.
From Strategy to Execution
Immediately. Populate coordinate two for the twenty roles with the heaviest counterparty contact, not the twenty most senior. Two hours per role with the person who holds it produces a defensible number.
Over the next two to four quarters. Change the instruments, not the exhortations. The schedule gains its second column. A register of preliminary instruments records who issued what, to whom, and against what value. Roles above the factor receive standing correspondence rules and a named reviewer for outbound assurances above a set value.
Over the longer term. The enterprise stops denominating its control environment exclusively in expenditure. Commercial conduct becomes a competence with a defined standard, taught to the people who generate the evidence.
Signals to Monitor
Watch the proportion of engagements in which a supplier begins work before an executed contract exists, as a rate rather than exceptions. Watch how long counterparty correspondence asserting an understanding the enterprise does not hold sits uncorrected. And watch how internal audit frames its findings: a run of items described as documentation lapses is usually a run of authority breaches nobody has the vocabulary to name.
Questions for the Leadership Team
- What is the largest sum any employee could cause this enterprise to owe today while inside their approval limit, and who computed it?
- How many instruments described as non-binding did we issue last year, who signed them, and what was the value of the work they named?
- In how many supplier engagements did work begin before an executed contract existed, and how long was the gap?
- When a counterparty last asserted an understanding we did not hold, how long did we take to correct it in writing, and who owns that response?
- Which roles hold unaccompanied commercial contact with counterparties, and what is the exposure ceiling of the largest?
- What did our last three commercial disputes turn on — a clause we negotiated, or conduct nobody knew we had authorised?
Closing Perspective
A board that signs a delegation schedule believes it has drawn the enterprise's commercial boundary. It has drawn one side, in a unit it controls. The other is redrawn every working day, in ordinary correspondence, by people never told they are drawing it.
The choice now sitting with the leadership team is whether to keep operating a control system calibrated in a currency its exposure is not denominated in. Putting the second number beside the first is neither difficult nor expensive, only uncomfortable: for most enterprises it will be the larger, and once written down nobody can claim they were governing obligation when they were only ever governing spend.
About EraNorth Insights
EraNorth Insights publishes practical analysis on strategy, projects, operations, transformation and decision intelligence for professional and organisational use. About EraNorth.