Authority Ends at the Boundary — Coalition Building Must Begin There

Leadership · No. 22
Originally written November 2005 | Re-edited and re-rendered for the WhereBeyond Edition, 2025

A coalition is not a list of supporters; it is a negotiated system of commitments strong enough to survive the first costly decision.

Executive Summary

A familiar failure occurs in large change programmes. The chief executive approves the objective, the programme board approves the plan, and every function sends a representative to the launch meeting. Three months later, the work stalls at exactly the points where one part of the organisation must incur cost so that another can receive benefit. Attendance remains high. Support remains verbal. Decisions remain absent.

The mistake is to treat coalition building as stakeholder communication. Communication can create awareness and reduce misunderstanding. It cannot by itself reconcile competing incentives, transfer decision rights or make a functional leader accept a local disadvantage for an enterprise result.

This methodology treats a coalition as an operating arrangement rather than a collection of sympathetic people. It is designed for the federated and matrixed organisation of the present period: business divisions with their own profit responsibilities, central functions with policy authority but limited operational control, shared-service initiatives, enterprise systems, outsourced providers and programme structures that must work across all of them.

The method has six phases:

1. Mandate the shared outcome.

1. Map the real boundaries of authority and consequence.

1. Construct the exchange that makes participation rational.

1. Negotiate the coalition charter and its decision rights.

1. Activate commitment through early consequential decisions.

1. Maintain, repair and eventually dissolve the coalition.

Each phase produces a specific artefact and ends with a decision gate. The method is deliberately practical. It identifies the roles required, the evidence to gather, the conversations to hold, the tests to apply and the warning signs that indicate a coalition exists only on paper.

The central discipline is simple: never confuse agreement with commitment. Agreement is a view. Commitment is a view attached to an owner, a contribution, a decision right and a consequence.

Why formal sponsorship is insufficient

Across enterprise change, the organisation chart creates a misleading picture of power. It shows who can instruct whom within a reporting line. It does not show who can delay a data definition, retain a local process, withhold a subject-matter expert, protect a budget, influence a works council, reinterpret a policy or make implementation so difficult that a formally approved decision becomes impractical.

Cross-boundary change therefore has three kinds of authority.

• Hierarchical authority can direct work within a line.

• Expert authority determines what is technically, legally or operationally credible.

• Consequential authority belongs to those who can impose, absorb or avoid the practical cost of change.

A senior sponsor may possess the first and still lack the other two. A programme manager may have a mandate to coordinate and no authority to compel. A local finance director may be two levels below the sponsor and still control the budget line on which the promised saving depends.

This is why the familiar stakeholder plan often disappoints. It records interest and influence, assigns messages and schedules contact. It does not answer the harder questions: what must each party give up, what do they receive, which decisions can they make, which can they block, and what happens when interests diverge?

A coalition is not a list of supporters; it is a negotiated system of commitments strong enough to survive the first costly decision.

The conditions for using the method

The method is appropriate when no single executive can deliver the outcome through a direct chain of command. Typical situations include an enterprise resource planning implementation spanning business units, the creation of a shared service, a merger integration, a common procurement policy, a regulatory response involving several functions, or a major outsourcing transition.

It is not necessary for every cross-functional task. Use a simpler working group when the outcome is narrow, authority is clear and the parties bear roughly equal costs and benefits. A full coalition is warranted when at least three of the following conditions hold:

• two or more leaders control essential resources;

• costs and benefits fall in different parts of the organisation;

• local performance measures conflict with the enterprise objective;

• important knowledge sits outside the sponsor’s reporting line;

• one party can delay the outcome without formally rejecting it;

• external suppliers, employee representatives or joint ventures affect delivery;

• the outcome requires sustained behaviour after the programme closes.

Before beginning, the commissioning sponsor should accept one uncomfortable premise: coalition members are not being asked merely to comply. They are being asked to make and keep bargains. If the sponsor will not allow those bargains to shape scope, sequence or resource allocation, the exercise is consultation, not coalition building.

The operating roles

Coalitions fail when every participant is described as a stakeholder and nobody is assigned the work of maintaining the system between them. Six roles are required. One person may hold more than one role in a smaller initiative, but the responsibilities must remain distinct.

	Role	Primary responsibility	Authority required	Failure if absent
	Commissioning sponsor	Defines the enterprise outcome and arbitrates beyond the coalition’s delegated authority	Can commit enterprise priority and settle escalated conflicts	Coalition negotiates endlessly without a final court
	Coalition lead	Designs and runs the method; maintains commitments and repairs fractures	Direct access to principals and freedom to surface conflict	Meetings occur, but the coalition itself is unmanaged
	Boundary principals	Commit their functions, divisions or partners	Can allocate people, budget and local decision rights	Representatives agree to actions their organisations will not honour
	Operational envoys	Test feasibility and carry operational knowledge between boundaries	Credibility with practitioners and access to evidence	Senior agreement rests on implausible assumptions
	Evidence steward	Maintains facts, baselines, assumptions and decision records	Access to management information and finance records	Interests argue through incompatible versions of reality
	Constructive challenger	Tests whether agreement is genuine and voices suppressed objections	Permission to question the sponsor and majority	Apparent unanimity conceals unresolved resistance
::table
The coalition lead is not a ceremonial chair. The role requires political judgement, disciplined preparation and the ability to distinguish a personal objection from a structural conflict. It should not automatically be assigned to the most senior person or to the programme office. Seniority can inhibit candour; administration can crowd out negotiation.
## Phase One: Mandate the shared outcome
The first phase prevents the coalition from forming around a slogan. “Standardise procurement”, “integrate the functions” or “put the customer first” may be worthy intentions, but they are too elastic to govern trade-offs.
### Inputs
–> the approved strategic or programme objective;
–> the investment case and principal assumptions;
–> known constraints, including statutory obligations, contractual commitments and budget limits;
–> the sponsor’s statement of why action is required now.
### Work
The coalition lead holds a short mandate session with the commissioning sponsor. The session converts the objective into four statements.
–1. Outcome: What observable condition must exist when the coalition has succeeded?
–2. Boundary: What is inside the coalition’s remit, and what remains outside it?
–3. Non-negotiables: Which constraints are fixed, and who fixed them?
–4. Negotiables: Which elements may change if the coalition discovers a better route?
A strong outcome statement describes operational reality, not activity. “All six divisions use one approved supplier catalogue for defined indirect categories, with local exceptions authorised through one process” is testable. “Implement strategic sourcing” is not.
The sponsor must also name the enterprise consequence of failure. Without it, local leaders will reasonably treat the coalition as one priority among many.
### Output: the Mandate Note
The Mandate Note is one page. It contains the outcome, scope boundary, non-negotiables, negotiables, time horizon, failure consequence and sponsor’s retained decisions. Its brevity matters. A long charter at this point usually disguises ambiguity rather than resolving it.
### Gate: Is the problem genuinely shared?
Proceed only if the proposed members recognise the consequence, even if they disagree about the remedy. If several parties believe the problem belongs entirely to the centre, do not launch the coalition. Return to the sponsor and either alter the mandate or accept that formal direction will be required.
::callout
The first test of a coalition is not whether people support the solution. It is whether they recognise a consequence that none of them can manage alone.

Phase Two: Map authority, interests and consequences

Traditional influence maps place names in boxes. This phase maps the machinery beneath the names.

Inputs

• the Mandate Note;

• organisation charts and committee terms of reference;

• budget ownership and performance measures;

• relevant contracts, service agreements and policy authorities;

• initial interviews with boundary principals and operational envoys.

Work

For every boundary, identify five elements.

1. Contribution: What resource, permission, knowledge or behaviour must this party provide?

1. Benefit: What does the party gain if the outcome succeeds?

1. Burden: What cost, risk, loss of discretion or disruption must it absorb?

1. Authority: Which required decisions can it make, and which lie elsewhere?

1. Exposure: What can happen to the outcome if the party delays, withdraws or complies minimally?

Do not ask only what a leader says. Examine measures and budgets. A division president who endorses consolidation while being rewarded for local operating margin has a structural reason to resist costs that benefit the group. A central function that promises efficiency but charges implementation costs back to the divisions may have created opposition before the first workshop.

The coalition lead then distinguishes three positions:

• aligned: benefit and burden are broadly balanced;

• negotiable: imbalance exists but can be addressed through sequence, resource, recognition or reciprocal commitment;

• structurally opposed: the requested contribution damages a party’s legitimate obligations or measures without adequate remedy.

Structural opposition is not bad attitude. Treating it as such drives the conflict underground.

Output: the Boundary Map

The Boundary Map is a working table showing each party’s contribution, benefit, burden, authority, exposure and current position. A second page shows critical dependencies between parties. It should be reviewed privately with each principal before it is discussed collectively; otherwise participants spend the first meeting correcting their portrayal in public.

Gate: Is there a feasible coalition?

Proceed if all indispensable parties are aligned or negotiable. If an indispensable party is structurally opposed, the sponsor must change the mandate, alter incentives, provide compensation or use formal authority. Facilitation cannot solve an impossible bargain.

Phase Three: Construct the exchange

Coalitions endure when each member can explain why participation is rational to the people they represent. Moral appeals to “the good of the enterprise” are rarely sufficient where budgets, service levels and careers are exposed.

The exchange is not a payment. It is the total arrangement of burdens, benefits, protections and reciprocities.

Inputs

• the Boundary Map;

• quantified costs, benefits and capacity demands;

• the sequence of proposed changes;

• known local obligations and peak operating periods.

Work

The coalition lead develops an Exchange Ledger with each principal. For every required contribution, it records what the coalition offers in return. The available instruments are broader than money.

• Sequence: A division accepts the standard process after its year-end close rather than during it.

• Capacity: The programme funds temporary backfill for local experts.

• Protection: Service levels are guaranteed during transition, with an agreed remedy if they fall.

• Reciprocity: The central function simplifies a control that creates local work in return for adoption of the common policy.

• Voice: The party receives a defined decision right over design elements that affect its operations.

• Recognition: Enterprise contribution is reflected in objectives and performance review.

• Evidence: A pilot or controlled trial reduces uncertainty before full commitment.

The exchange must not purchase consent by surrendering the outcome. Local exceptions may be legitimate, but every exception has a cost in complexity. Record that cost and identify who accepts it.

Output: the Exchange Ledger

The ledger lists each contribution, burden, offered return, owner, timing and unresolved condition. It is confidential during construction and becomes a coalition record only after the principals agree what may be shared.

Gate: Is the exchange balanced enough to negotiate?

The test is not perfect equality. Some parties will carry more burden because the enterprise requires it. The test is whether each principal can defend the arrangement as legitimate, and whether hidden concessions have been avoided. If agreement depends on promises outside the sponsor’s authority, stop and secure that authority first.

Phase Four: Negotiate the Coalition Charter

Only now should the group meet to form the coalition. Beginning with a large workshop before the mandate, boundary and exchange work is complete produces theatre: public enthusiasm followed by private retreat.

Inputs

• the Mandate Note;

• agreed portions of the Boundary Map;

• the Exchange Ledger;

• a draft decision-rights schedule;

• a first version of the commitment register.

Work

The formation meeting addresses decisions, not presentations. The coalition lead circulates evidence in advance and structures the meeting around five questions.

1. What outcome are we jointly committing to?

1. What will each boundary contribute?

1. Which burdens and protections have been agreed?

1. Which decisions belong to the coalition, to individual principals and to the sponsor?

1. How will we recognise and repair a broken commitment?

Decision rights should be explicit. A conventional responsibility chart may clarify tasks, but it is insufficient for contested choices. The charter must identify who proposes, who supplies evidence, who must concur, who decides and who may appeal.

Use concurrence sparingly. Requiring every function to concur gives each an unacknowledged veto. Where a decision affects several parties, identify one accountable decider after proper consultation. Reserve unanimity for changes to the coalition’s mandate or exchange.

The constructive challenger tests each commitment by asking: “What will your organisation observe you doing differently next month?” If the answer is another meeting, the commitment is not yet operational.

Outputs

The Coalition Charter records purpose, membership, decision rights, escalation, meeting discipline, information rules and dissolution conditions.

The Commitment Register records a small number of consequential commitments with owner, due date, evidence and dependent parties.

The Decision Record captures what was decided, the evidence used, objections raised, conditions attached and review date.

Gate: Has authority been committed?

Each boundary principal must confirm that they possess the authority they are exercising. If a delegate needs later approval, the commitment remains provisional and is labelled accordingly. The coalition is not active until all indispensable contributions have an authorised owner.

Phase Five: Activate through consequential decisions

A charter creates potential energy. The coalition becomes real only when members make decisions that cost something: allocate scarce people, release budget, change a local measure, close an exception or accept a constraint.

Inputs

• the Coalition Charter;

• the Commitment Register;

• the first 90-day delivery plan;

• baseline measures for the shared outcome.

Work

Select two or three early decisions that are consequential, visible and reversible only with effort. Avoid symbolic quick wins that prove merely that people can cooperate when nothing important is at stake.

The coalition lead runs a monthly cycle.

1. Prepare: Evidence steward issues a concise pack showing outcome measures, commitments due, emerging conflicts and decisions required.

1. Test: Operational envoys confirm what is happening at the boundary, not what the central plan assumes.

1. Decide: Principals make the choices reserved to them in the charter.

1. Commit: The register records owner, contribution, evidence and consequence.

1. Verify: At the next cycle, evidence confirms whether the commitment changed behaviour.

1. Repair: Missed commitments trigger diagnosis and remedy, not automatic blame.

A missed commitment can arise from unwillingness, lack of authority, changed conditions, poor feasibility or a dependent party’s failure. The remedy should match the cause. Repeated unexplained delay is treated as withdrawal from the coalition and escalated.

Output: the Coalition Scorecard

The scorecard should fit on one page. It shows no more than:

• progress against the shared outcome;

• contribution delivered by each boundary;

• overdue commitments;

• decisions required before the next cycle;

• unresolved exchange imbalances;

• confidence in the coalition’s ability to deliver.

Do not reduce coalition health to red, amber and green alone. A green status can conceal a commitment that has not yet been tested. Add a short evidence statement and the next consequential decision.

Gate: Has the coalition survived cost?

After 90 days, ask whether each indispensable party has made at least one consequential contribution and whether the shared outcome has moved. If not, the coalition remains nominal. The sponsor must renew the exchange, exercise authority or stop the initiative.

Phase Six: Maintain, repair and dissolve

Coalitions are temporary systems. They must adapt as people, incentives and evidence change, and they must end when their work is complete. Permanent coalition meetings become another layer of governance detached from consequence.

Maintenance rhythm

Quarterly, the coalition lead revisits the Boundary Map and Exchange Ledger. This is essential after reorganisations, budget rounds, supplier changes or leadership appointments. A new principal inherits neither the predecessor’s understanding nor their informal bargains automatically.

At each review, ask:

• Has the shared outcome changed?

• Have burdens moved between boundaries?

• Are benefits arriving where expected?

• Do principals still hold the required authority?

• Which commitments have become routine operations?

• Which unresolved disputes now require sponsor judgement?

Fracture diagnosis

When cooperation deteriorates, classify the fracture before acting.

	Fracture	Observable sign	Likely cause	Appropriate repair
	Mandate fracture	Members pursue different outcomes	Ambiguous or changed purpose	Sponsor restates or revises mandate
	Exchange fracture	One party carries accumulating burden	Benefits or protections failed to materialise	Renegotiate sequence, resource or reciprocity
	Authority fracture	Decisions repeatedly return for approval	Principals lack delegated rights	Replace representatives or change delegation
	Evidence fracture	Parties dispute the facts	Baselines or definitions diverged	Evidence steward rebuilds common record
	Trust fracture	Information is withheld or surprises recur	Commitments were broken without acknowledgement	Explicit breach conversation and new verification
	Capacity fracture	Willing parties cannot deliver	Resource assumptions were unrealistic	Replan, backfill or narrow scope
::table
Do not use team-building activity to treat a structural fracture. Trust may matter, but trust cannot compensate indefinitely for an unfair exchange or absent authority.
### Dissolution
The coalition should dissolve when the shared outcome is embedded in ordinary management, when the mandate is withdrawn, or when evidence shows the bargain is no longer viable.
The closure record identifies:
–> outcomes achieved and not achieved;
–> commitments transferred into operational management;
–> decisions and assumptions still open;
–> lessons about boundaries, authority and exchange;
–> the owner of any continuing measure.
Dissolution is a sign of completion, not failure. The aim is to build enough shared authority for a purpose, then return accountability to the organisation’s enduring structures.
## Worked example: a shared procurement coalition
A composite enterprise in late 2005 had six operating divisions, 18 sites and £72 million of annual indirect expenditure. Central procurement estimated that common catalogues, consolidated volume and fewer suppliers could save £5.8 million a year. The executive committee approved the objective. Yet after four months, only two divisions had supplied reliable spend data and none had agreed to surrender local supplier lists.
The initial diagnosis described “resistance to standardisation”. The Boundary Map showed a more useful picture.
–> Central procurement received the saving in its enterprise case but carried little implementation cost.
–> Division finance directors carried data-cleansing work and feared disruption to year-end controls.
–> Site managers relied on local suppliers for urgent maintenance and were measured on plant availability.
–> Information systems staff were already committed to a general-ledger consolidation and could not support another major interface change.
–> The accounts-payable shared service would receive fewer invoices eventually but faced a surge of supplier amendments first.
The problem was not insufficient communication. The proposed exchange asked every other boundary to bear immediate cost for a benefit credited to the centre.
The coalition was rebuilt through the six phases.
The Mandate Note narrowed the first outcome to four categories representing £31 million of spend. Emergency maintenance was excluded temporarily. The Boundary Map identified nine indispensable principals rather than the 34 names on the original stakeholder list.
The Exchange Ledger funded six temporary data analysts for ten weeks, delayed one division until after year-end close, guaranteed a local emergency route and assigned half of verified first-year savings to division budgets. In return, each division committed a finance manager, a procurement lead and authority to close unauthorised supplier accounts.
The Coalition Charter gave category decisions to the procurement director after concurrence from the affected division finance directors. Site-level safety exceptions remained local but required written evidence. The sponsor retained decisions on disputes above £250,000.
The first consequential test was the office-supplies category. The coalition closed 146 duplicate supplier accounts, introduced one catalogue and moved 83 per cent of eligible orders within eight weeks. Average order cost fell from £38.40 to £31.10. More important than the early £210,000 annualised saving, three divisions changed their local approval routines and released named staff to the next category.
Not everything worked. A facilities category stalled because the proposed supplier could not meet two remote sites’ response requirements. The coalition did not label the sites resistant. It reclassified the issue as an exchange and evidence fracture, retained two regional suppliers, recorded the £170,000 reduction in forecast savings and protected the operational outcome.
After twelve months, verified annualised savings were £4.4 million rather than £5.8 million. Four categories were standardised, invoice exceptions had fallen by 23 per cent, and the divisions accepted the continuing governance into the annual planning cycle. The lower number was more credible because the coalition had preserved the conditions that produced it.
The turning point was not the launch workshop. It was the moment the centre paid for data work, the divisions changed their approval routines and the coalition accepted a lower saving to protect plant availability. The bargain became real because every boundary both gave and received something consequential.
## Practical deployment plan
A practitioner can establish the method over eight to ten weeks before the main delivery cycle.
::table
	Week	Activity	Primary role	Required output
	1	Sponsor mandate session	Commissioning sponsor and coalition lead	Mandate Note
	2–3	Principal interviews and evidence review	Coalition lead and evidence steward	Draft Boundary Map
	4	Operational validation	Operational envoys	Confirmed contributions, burdens and exposures
	5–6	Bilateral exchange design	Coalition lead and boundary principals	Exchange Ledger
	7	Decision-rights design and challenge	Sponsor, coalition lead and challenger	Draft charter and commitment register
	8	Coalition formation meeting	All principals	Authorised charter
	9–10	First consequential decisions	Boundary principals	Active commitments and baseline scorecard
::table
The work is not linear in the sense of being irreversible. New evidence may send the coalition back to the boundary map or exchange. The gates prevent this iteration from becoming drift.
## Common mistakes and their correction
–> Starting with a large workshop. Begin with bilateral diagnosis. Public meetings are poor places to discover private burdens.
–> Inviting every interested person into the coalition. Keep principals to the smallest set that controls indispensable contributions. Use wider consultation around it.
–> Selecting representatives for availability. Require authority and credibility. A diligent delegate without decision rights creates false progress.
–> Treating resistance as a personality defect. Examine measures, budgets, obligations and exposure before judging motive.
–> Offering only enterprise benefit. Construct a defensible exchange at each boundary.
–> Leaving decision rights implicit. Record who proposes, provides evidence, concurs, decides and appeals.
–> Counting attendance as health. Test the coalition through costly, observable commitments.
–> Escalating every disagreement. Reserve the sponsor for disputes outside delegated authority; otherwise the coalition never develops its own capacity.
–> Keeping the coalition alive indefinitely. Transfer routine accountability and dissolve the temporary structure when its purpose is complete.
## Final test
A coalition is working when members can disagree without withdrawing, when local burdens are visible rather than moralised, when decisions occur at the agreed level, and when a broken commitment produces repair or consequence.
The methodology does not eliminate politics. It makes politics discussable in operational terms: authority, exchange, evidence and commitment. That is the difference between a coalition that looks impressive in a steering pack and one that can move an organisation across the boundaries its chart was never designed to cross.