Alliance contracting: choosing partners before price and keeping open-book costs disciplined

Some projects cannot be fully defined before work starts. Upgrading a treatment plant while it keeps operating, rebuilding rail infrastructure under live traffic, or delivering a complex facility with unknown ground conditions all involve uncertainty that no fixed-price tender can price honestly. Conventional contracts push that uncertainty onto contractors, who either price it heavily or fight about it later through variations and claims.

Alliance contracting takes a different approach. The owner and its delivery partners form a single integrated team, share decisions and share the financial outcome against an agreed target cost, with costs reimbursed on an open-book basis. Alliances have been widely used for major infrastructure in Australia. They can produce excellent results on the right projects. They can also produce expensive ones, when the wrong partners are chosen, when targets are soft, or when open-book costs are not kept disciplined.

This article explains how alliances work, when they suit a project, how partners can be selected on observed capability rather than written claims, how the commercial model works, and how owners keep open-book costs and targets honest. It also covers what businesses joining alliances need to be ready for. It is general information. National and state alliance contracting guidelines in Australia set out detailed practice, which should be consulted for specific projects.

How an alliance works

In a project alliance, the owner and one or more non-owner participants, such as a designer and a constructor, sign a single agreement. Common features include:

• An alliance leadership team of senior representatives from each party, which sets direction and makes key decisions, usually unanimously.
• An alliance management team that runs the project day to day, drawn from all parties.
• An integrated project team, with people from all parties working together, often co-located.
• Best-for-project decisions, taken in the interest of the project rather than any one party.
• No blame and no dispute provisions, under which parties generally agree not to sue each other except in limited cases such as wilful default.
• Shared risk and reward against agreed cost and performance targets.

The owner is a full participant, not a distant client. That is both a strength, because the owner’s knowledge and decisions are inside the team, and a demand, because the owner must provide capable, senior people for the life of the project.

When an alliance suits a project

Alliances suit projects with:

• High uncertainty that cannot be priced efficiently by a contractor.
• Complex risks that are best managed jointly.
• Scope that cannot be fully defined at the outset.
• Many interfaces, such as live operations, other contractors or stakeholders.
• A need for owner involvement in ongoing decisions.
• Time pressure that rewards starting before design is complete.

They do not suit well-defined, routine work, where competitive fixed-price or schedule-of-rates contracts give clearer value. Collaboration does not depend on an alliance; many collaborative practices can be built into ordinary contracts, as the collaboration is a practice, not a contract article explains. The matching the contract to the work article covers choosing contract structures more broadly.

The commercial model

Alliances commonly pay non-owner participants under three limbs:

• Limb 1: reimbursable costs. The actual direct costs of the work, paid on an open-book basis and subject to audit.
• Limb 2: corporate overhead and profit. A fee, usually agreed as a percentage or lump sum, covering the participant’s overheads and normal profit.
• Limb 3: gainshare and painshare. A sharing of the difference between the actual cost and the target outturn cost, plus rewards or penalties for performance in key result areas such as safety, quality, environment, community and schedule.

The target outturn cost is the estimate of the total cost of delivering the project, developed jointly before the main works are committed. If actual costs come in below it, the saving is shared between the owner and participants; if above, the overrun is shared. Participants’ painshare is commonly capped, often so they risk their profit and overhead but not their direct costs.

This model aligns incentives: everyone gains from lower cost and better performance. It also depends entirely on the target being realistic and the costs being genuine.

The phases of an alliance

Alliances usually move through distinct phases, each with its own decisions:

1. Selection: choosing the non-owner participants through staged evaluation.
2. Development: the alliance team develops the design, identifies and prices risks, and builds the target outturn cost and performance targets. Participants are usually paid for this work on a reimbursable basis.
3. Commitment: the owner reviews the target and business case and decides whether to proceed. A critical protection is that the owner can choose not to proceed, or to proceed another way, if the target does not offer value for money.
4. Implementation: the alliance delivers the works, managing risks and costs against the target, with regular reporting to the leadership team.
5. Completion and close-out: final costs are audited, gainshare or painshare is calculated, defects are resolved and lessons are recorded.

Treating the commitment decision as a formality is one of the biggest risks to value. By the end of development, the owner and team have invested time and goodwill, and there is strong pressure to proceed. A disciplined review against independent estimates and the business case protects the owner from committing to a target that is too high.

Choosing partners before price

A conventional tender asks which supplier offers the strongest compliant proposal at the best price. Alliance selection must also ask which group of people can work credibly inside an integrated governance and commercial system when things get difficult. Some of the most important capabilities are behavioural:

• Can senior leaders challenge each other constructively?
• Can the team make sound decisions with incomplete information?
• Will commercial staff share cost data openly rather than tactically?
• Can technical leaders combine owner and contractor knowledge?
• Can disagreements be resolved without retreating to contractual positions?

These are hard to judge from written claims. Australian alliance practice therefore uses staged selection, moving from expressions of interest to proposals, a shortlist and an interactive development phase. During that phase, shortlisted teams work with the owner in workshops to develop the solution, analyse risks and test the commercial framework, and the owner observes how they behave.

A strong selection process assesses six dimensions, with the evidence for each defined before interaction begins:

Common mistakes include assessing personality rather than observable behaviours, being impressed by polished workshop performances that may not last, and letting the owner become emotionally committed to one team before final evaluation.

Probity in interactive selection

Interactive development creates probity risks. When shortlisted teams work closely with the owner, information must be controlled so that one team does not receive another’s ideas. Owner staff need clear protocols, sessions need records, and commercial concessions should not be made informally in workshops. Owners should also guard against an incumbent or favoured team gaining hidden advantages.

Keeping price competition

Selecting on capability does not mean ignoring price. Many owners now introduce competition on commercial terms, for example by comparing participants’ fees and key rates, or by having two shortlisted teams develop target costs in parallel before selecting one. More competition improves price tension but increases bid costs and time, so the approach should suit the project’s size and uncertainty.

Keeping open-book costs disciplined

Alliances fail commercially when targets are soft and costs drift. Owners can protect value in several ways:

• Independent estimating. Commission an independent estimate of the target outturn cost and reconcile differences line by line before agreeing the target.
• Clear scope at target. Record exactly what the target covers, so later changes can be separated from normal development of the design.
• Strict target adjustment rules. Adjust the target only for genuine owner-directed scope changes, not for risks the team should have managed or estimates that proved low.
• Clear cost definitions. Define what can be claimed as a reimbursable cost and what belongs in the fee, so overheads and profit are not recovered twice.
• Regular audit. Audit reimbursable costs throughout the project, not just at the end, including timesheets, plant rates and subcontract payments.
• Transparent forecasting. Require regular forecasts of final cost against target, with explanations of movements.
• Meaningful key result areas. Set performance measures that reflect real project outcomes, with thresholds that require genuine performance, not routine compliance.
• Benchmarking. Compare rates, quantities and productivity with other projects.
• Capable owner representatives. The owner’s people on the leadership and management teams need commercial expertise and the authority to challenge.

An alliance can also run into trouble even with good intentions. Agreements should include procedures for leadership team deadlock, for replacing people whose behaviour undermines the alliance, and for termination where necessary.

Sustaining the relationship

The behaviours observed during selection have to last for years. Alliance leadership teams maintain them deliberately: by agreeing principles and behaviours at the start, inducting new team members, holding regular health checks of the relationship, addressing poor behaviour quickly regardless of which party the person comes from, and celebrating best-for-project decisions that cost an individual party something. Pressure tests the culture. When costs rise, schedules slip or something goes wrong on site, the temptation to revert to blame and contractual positions is strong, and that is exactly when leaders need to model the agreed behaviours.

For businesses joining alliances

Designers, constructors and specialist contractors joining alliances should prepare for:

• Open-book cost systems: accurate job costing, timesheets and records that can withstand audit.
• Seconding capable people for long periods, including senior leaders.
• Cultural fit: staff who can work transparently and share problems early.
• Clear subcontract arrangements: subcontracts within alliances are often conventional contracts, and their terms need to fit the alliance’s objectives.
• Commercial discipline: understanding how the fee, target and gainshare interact, and the realistic chance of painshare.

A worked example

This is an illustrative example. A water utility needs to upgrade a treatment plant while keeping it in service. Ground conditions are uncertain, the plant’s existing equipment is poorly documented and outages must be scheduled around demand. The utility chooses an alliance.

Selection. After expressions of interest and written proposals, two teams of designers and constructors are shortlisted for a paid development phase. Over several workshops, evaluators observe how each team handles a simulated outage problem and a disagreement about risk. One team’s leaders defer to whoever is most senior; the other team’s engineers challenge their own commercial manager’s assumptions constructively. The second team is selected, with fees and key rates also compared.

Target. The alliance develops a target outturn cost of $86 million. An independent estimate comes in at $82 million. Line-by-line reconciliation finds duplicated contingency and conservative productivity rates, and the target is agreed at $84 million, with scope and assumptions recorded.

Delivery. Costs are audited quarterly. An audit finds $0.4 million of overhead costs claimed as reimbursable costs, which are corrected. Two design changes requested by the utility adjust the target under the agreed rules. The project finishes at $81 million against the adjusted target of $84 million, and safety and water quality measures are met.

Outcome. The $3 million underrun is shared equally, so participants receive $1.5 million in gainshare. The utility records what the selection evidence predicted and how it compared with the team’s actual performance, to improve future selections.

Applying this in an Australian context

• Use alliances only where uncertainty and complexity justify them.
• Select partners on observed capability, with evidence defined in advance.
• Manage probity carefully during interactive selection.
• Introduce appropriate price competition.
• Validate the target with independent estimating.
• Define costs clearly and audit regularly.
• Set meaningful key result areas.
• Staff the owner’s side with capable, senior people.

Where alliances go wrong

• Alliances for routine, well-defined work.
• Selecting on presentation skills rather than observed behaviour.
• Soft targets accepted without independent validation.
• Target adjustments for risks the team should have managed.
• Overheads recovered twice through loose cost definitions.
• Key result areas that reward routine compliance.
• A weak owner team unable to challenge.

Questions for the leadership team

• Is this project uncertain and complex enough to justify an alliance?
• Which capabilities can we only judge by observing teams at work?
• How will we keep competition and probity in selection?
• How do we know the target outturn cost is realistic?
• What rules govern changes to the target?
• Do we have the people to be an effective owner participant?

Bringing it together

Alliance contracting can deliver uncertain, complex projects well by integrating the owner and its partners, sharing decisions and sharing the financial outcome. Its success depends on two disciplines. First, choose partners before price, using staged selection and observed behaviour against defined capability dimensions, while protecting probity and keeping appropriate competition. Second, keep open-book costs honest, with independently validated targets, strict adjustment rules, clear cost definitions, regular audit, meaningful performance measures and a capable owner team. With both in place, alliances can achieve outcomes that conventional contracts struggle to reach.

---

Source: KEVOS editorial notes, drawing on earlier KEVOS corporate articles on alliance procurement, interactive partner selection and alliance commercial frameworks, together with general knowledge of Australian alliance contracting practice. The worked example is illustrative. This article is general information; consult current national and state alliance contracting guidelines for specific projects.