When compliance depends on supervision you have bought a cost you must keep paying; when it depends on the counterparty's own balance sheet you have bought an outcome.
A delivery is slipping. Quality is drifting. The instinct in most organisations is almost automatic: more oversight. Additional reporting lines, a weekly review that becomes twice weekly, an inspector on site, a governance forum with an escalation path. Each step is defensible alone, and together they produce a recurring cost that generates information rather than performance.
The uncomfortable arithmetic is that supervision scales with scope while its effect decays with distance. You cannot watch everything, you learn about failures after they occur, and the counterparty always knows more about their own work than you can discover from outside. Contracts allocate blame after the fact; reviews describe what has already happened. Neither changes the probability that the work is done well when nobody is looking.
There is a different question available, and senior leaders ask it too rarely. Can the counterparty's economics be arranged so that finishing on time, to standard, is what they would choose purely in their own interest? Where the answer is yes, supervision becomes a check rather than a control, and the cost of assurance falls as the quality of delivery rises.
The Strategic Context
Every commercial relationship carries an information asymmetry. The party doing the work knows its true state; the party paying sees a report. Governance narrows that gap imperfectly and expensively, and the thinner the margin on the work, the greater the counterparty's incentive to manage the report rather than the work — the report being cheaper to improve.
Penalties are the usual answer and weaker than they appear. A damages clause is priced into the bid before it is ever triggered — the counterparty has sold you insurance and charged you for it. It is invoked late, argued over, and often settled at a discount. Penalties transfer consequence; they do not change behaviour during the work, when behaviour is still available to be changed.
Incentive design is a different instrument. It asks what the counterparty holds, what that holding is worth, and what happens to its value if the outcome you want fails to arrive. If nothing happens to it, you have a supervision arrangement whatever the contract calls itself.
None of this substitutes for the prior decision of whether the work belongs outside the organisation at all. [Related article: Is Your Make-or-Buy a Cost Decision or a Capability Decision?] takes up that question. What follows assumes it is settled and the counterparty exists.
The Supervision Reflex
The first misreading is that more oversight buys more performance. Beyond a threshold it buys documentation and degrades the relationship that produces early warning. The people best placed to tell you a project is in trouble stop doing so when telling you triggers an audit.
The second treats penalties as incentives. An incentive changes what someone wants to do; a penalty changes what they pay if caught. Sophisticated counterparties price the second and manage the first.
The third treats alignment as a cultural matter. Partnership language, joint workshops and shared values are worth having, and none survives the moment when one party's margin depends on the other's concession. Culture supports an aligned structure; it does not replace one.
The fourth is subtler. Executives design incentives for the counterparty organisation and forget that the work is done by individuals whose own incentives may point elsewhere. A site manager rewarded on monthly cost has different objectives from a firm rewarded on completion. An arrangement that aligns the entity but not the people executing it will disappoint, and the disappointment will look inexplicable from the boardroom.
Reframing the Issue
The reframe is to stop designing controls and start designing claims. A control is something you operate. A claim is something the counterparty holds — and if you construct it well, they defend it without being asked.
The governing question becomes concrete: what asset can this counterparty hold whose value depends on the outcome we want? Not on their effort, which is unobservable, nor on a metric, which can be managed, but on the outcome itself. Where such an asset exists and can be transferred, the counterparty stops being a fee-taker and becomes a residual claimant. The party with the best information about the work is now the party most exposed to its failure.
Payment in Kind: Making a Supplier an Owner of the Outcome
The clearest instance is deceptively simple. Pay part of the consideration in cash and part in a completed unit of the thing being built — a contractor takes a share of the finished units. The logic generalises: output rights from the plant they commission, capacity in the facility they build, a revenue share on the line they install.
The effect is structural rather than motivational. The counterparty now holds something whose value is impaired if the work is unfinished and impaired again if it is finished badly. Delay costs them realisation, defects reduce their own asset, and abandoning the site destroys their position. You have not persuaded them to care about quality; you have made carelessness expensive to them at the moment they would otherwise be tempted by it.
Note what is transferred and what is not. Payment in kind hands over a claim on output — negotiated and bounded. That is a different order of commitment from handing over the capability to produce, which is far harder to recover once gone. [Related article: What Did You Hand Your Contract Manufacturer That You Cannot Get Back?] deals with that second and more consequential transfer.
Four conditions determine whether it works. The unit must be liquid enough that the counterparty would rather hold it than not. Their share must be material against their cash margin, or it becomes a rounding item. Their claim must be robust — if your insolvency or a financier's prior security can defeat it, the incentive evaporates exactly when it is most needed, and this is where such arrangements most often fail. And the counterparty must be capitalised well enough to carry a non-cash asset; offering it to a thinly financed subcontractor is a wage cut dressed as partnership.
The failure modes are equally specific. Selective effort — perfecting the unit they will own and neglecting the rest — is answered by tying their share to the whole increment rather than an identified item. Valuation disputes are answered by fixing the basis before work starts. Adverse selection is the quiet one: if only counterparties who cannot obtain cash accept payment in kind, the mechanism has selected for weakness rather than alignment. Offer it against a fair cash alternative and watch who takes it.
Sequencing a Recovery: Cash Before Margin, Scope Before Ambition
Incentive design matters most when a programme has already stalled, because supervision is then at its least effective and every party is reassessing whether to stay. Three sequencing decisions carry most of the weight.
The first is authority. Stalled work usually has several people who can say no and nobody who can say yes. A single accountable decision-maker, with authority to commit and responsibility for the result, removes the delay a recovery can least afford. Delay is the mechanism by which stalled work becomes abandoned work.
The second is scope. The instinct is to restart everything at once and demonstrate momentum. The better move is to define the smallest coherent increment that discharges an existing obligation and complete it fully. Finishing something for a party who has already paid restores the flow of money and the willingness of everyone else to act. Narrow targets are not modest ambition; the purpose of the first is to show that a target can be met at all. This is the crisis form of a discipline that ought to be permanent — [Related article: One Thing Exceptionally, or Seven Things Adequately?] treats concentration as a standing portfolio choice, not an emergency measure.
The third is the objective function. In recovery, cash conversion outranks profitability, and leaders who cannot say so out loud will make a series of margin-preserving decisions that extend the crisis. Accepting a thinner return on the first increment is not weakness; it buys the credibility everything afterwards depends on.
The same logic governs exits. A counterparty or customer who wants out will ask for their money now, and paying it drains the cash the recovery runs on. Staged, completion-linked exit terms keep their interest attached to completion rather than against it. The party leaving becomes, briefly, one more party who wants the work finished.
When Restraint Substitutes for Incentive
Organisations frequently reach for contractual restraint where incentive design would serve better. Advisory arrangements are the common case: a business assembles experienced advisers, becomes anxious about what they know, and tries to bind them with a long restraint on working anywhere near the sector.
In Australia, restraint-of-trade provisions are subject to reasonableness tests and a lengthy restraint on an adviser is of doubtful enforceability. [FACT CHECK REQUIRED] This requires professional legal verification for your circumstances; ERANORTH is not a law firm or a financial adviser. The practical risk is not only that the clause fails, but that the organisation believes it is protected and designs nothing else. Confidentiality obligations are narrower and generally more defensible, addressing the actual concern — use of specific confidential information — rather than a person's right to work. [FACT CHECK REQUIRED]
The incentive alternative is more durable in any case. An adviser engaged against a defined mandate, with a few agreed objectives, a regular review of whether they were met, and compensation split between a modest cash component and a deferred outcome-linked one, has a continuing reason to stay useful and a real cost in walking away. Value comes from the clarity of the mandate and the review, not the length of the restraint. A restraint you cannot enforce is a governance illusion; an incentive you never have to enforce is a governance asset.
Decision Framework
Test any proposed arrangement against six questions before it is signed.
| Test | Diagnostic question | Failure signal |
|---|---|---|
| Information | Who knows more about this work, and how else would we find out? | Assurance rests on their reporting alone |
| Stake | What do they lose automatically if the outcome fails? | Nothing but the prospect of future work |
| Robustness | Can their claim be defeated by our insolvency or a prior security? | The claim is unsecured or unregistered |
| Proportionality | Is the stake material against their cash margin? | Small enough to be written off |
| Scope of reward | Does the payoff attach to the whole outcome or a part? | They can succeed while the increment fails |
| Reversibility | Can we unwind this if it behaves badly? | Exit requires their consent |
Two thresholds apply. If you cannot answer "what do they lose, automatically, if this fails?" in one sentence, you have designed supervision and labelled it alignment. And if the answer depends on a legal process to realise, discount it heavily: an incentive that needs litigation to bite is a penalty wearing better clothes.
From Strategy to Execution
Begin with a single troubled engagement. Map the counterparty's actual payoff — not the contract's rhetoric, but what they receive, when, and what they forfeit if the work fails. The honest answer is usually future work they may not want anyway. That gap is the finding, and it generalises.
The medium-term work is capability. Writing these instruments needs commercial and legal drafting skill, finance capability to value non-cash consideration, and security expertise to make a counterparty's claim robust. It also needs a record of what has been tried, because the useful knowledge is specific: which structures held, which were gamed, and how.
The long-term positioning is a shift in procurement doctrine. An organisation that habitually buys supervision accumulates overhead and adversarial relationships. One that designs alignment spends more effort before signature and far less afterwards, and becomes a counterparty capable suppliers want to work with — itself a sourcing advantage competitors find hard to match.
Signals to Monitor
Track assurance cost as a proportion of contract value and watch its direction. A rising line means the organisation is buying more control and getting less for it.
Watch where defects are discovered. Late discovery, at handover or in service, indicates quality managed for inspection rather than for use. Watch what happens between reporting cycles, the only period in which incentive and supervision give different answers.
Watch how bidders treat penalty clauses: if they price them, the clauses are insurance rather than deterrence. Watch counterparty balance sheets for the capacity to hold non-cash consideration, since an arrangement they cannot afford to accept is not available to you. And keep the legal position on restraint and security of claims under review. [FACT CHECK REQUIRED]
Questions for the Leadership Team
- For each of our three largest delivery relationships, what does the counterparty lose automatically — without litigation, negotiation or our intervention — if the outcome fails?
- What are we spending on assurance across the portfolio, and what would we spend if the counterparties' own interests carried more of that load?
- Where we have used contractual restraint, have we verified that it is enforceable, and what have we stopped designing because we believe it is?
- In our current recovery situations, who holds single accountable authority, and if the answer is a committee, what is that costing in elapsed time?
- If a counterparty read our incentive structure as a puzzle to be solved, where would they solve it?
Closing Perspective
The discipline described here is not cleverness with contracts. It is a change in the question leaders ask when performance disappoints. "How do we make sure they do it?" leads to cost, distance and documentation. "What would make them want to?" leads to a structure that works when the room is empty.
Two cautions belong with it. Mechanism design does not substitute for choosing counterparties well; a good structure with the wrong party fails more expensively. And it is no licence to stop looking. The best mechanisms are the ones you would still be comfortable with if the counterparty treated them as a puzzle — because a capable one will.
What separates a durable commercial relationship from a supervised one is whether both parties would still do the work the same way if nobody were watching. That is a design question, answered at signature rather than at escalation.
About the author
Kevin Jogin is Founder & Principal Advisor at EraNorth. Meet the Founder.