ECI, Alliances and Collaborative Contracting: The Delivery Models Built for Majors, and Where an SME Actually Fits

Most civil contractors learn tendering in one shape: the client designs the job, issues the drawings and specification, and asks for a price. You read it, you price it, you win or you do not. Every other model looks like a variation on that theme, and the language used to describe them — ECI, alliance, target cost, collaborative contracting — sounds like something that happens on billion-dollar projects and has nothing to do with a business running eight machines.

That was true for a long time. It is less true now. Water authorities run programme-style frameworks with contractors embedded from concept. Councils run early contractor involvement on jobs measured in single-digit millions. State agencies run incentivised construct-only contracts where a share of the savings comes back to you. And every alliance ever formed subcontracts a large amount of ordinary civil work to ordinary civil contractors, under terms shaped by the alliance above it.

The models are worth understanding for two reasons. The first is that when one appears, the tender does not behave the way you expect and the standard bid response scores badly. The second is that the terms flow downhill: if you subcontract into a collaborative job, the head contract’s mechanisms shape your package whether or not anyone explains them to you.

The spectrum, from construct-only to alliance

It helps to see these as points on a line rather than as separate species. The line runs from the client carries the design and you carry the build at one end, to everybody carries everything together at the other. What moves along the line is the answer to three questions: who does the design, when is the price fixed, and who wears the difference when it costs more than expected.

Model	Design	When the price is fixed	Who wears the overrun
Construct-only	Client’s	At tender, before you start	You, except for defined relief events
Design and construct	Yours	At tender, against a brief	You, including design adequacy
Incentivised construct-only	Client’s	At tender, plus a savings mechanism	You, but savings are shared
ECI (two-stage)	Joint, then client’s or yours	At the end of Stage 1	You, once Stage 2 is agreed
Managing contractor	Client’s consultants, you coordinate	Progressively, package by package	Mostly the client
Target cost / pain-gain	Varies	As a target, not a fixed sum	Shared to a formula, usually capped
Alliance	Joint, inside the alliance	As a target agreed by all parties	Shared, with limited or no recourse

None of these replaces the contract form. An ECI still lands on a standard-form contract for Stage 2, usually one of the forms covered in our guide to contract forms beyond construct-only. A target cost arrangement is a payment mechanism bolted onto a contract, not a contract in itself. The delivery model tells you how the work is bought; the contract form still tells you who carries what. Reading one and assuming the other is a reliable way to be surprised.

Why a client chooses a collaborative model

Understanding the client’s reason matters, because it tells you what they will actually score. A client does not move away from lump sum because they enjoy complexity. They do it when one of a small number of conditions applies.

• The scope genuinely cannot be defined yet. Renewal work inside an operating asset, remediation where the extent is unknown until you open it up, or a programme of works where the individual jobs are not yet identified.
• The risk is unpriceable at a sensible number. When a client asks for a fixed price on genuinely unknown ground conditions, they receive either a contingency-loaded bid or a bid that will become a claim. Sharing the risk is often cheaper than buying it.
• Construction knowledge is needed during design. A design produced without buildability input generates variations. Bringing a contractor in early is an attempt to buy that input before the drawings are frozen.
• Programme certainty outranks price certainty. Where an outage, a season, an event or a funding deadline is immovable, the client will trade cost certainty for the ability to start early and adjust.
• The relationship is ongoing. On a framework or panel, the client is buying a capability over years rather than a job, and behaviour matters more than the sharpest single price.

Read the tender documents for which of these is driving it. It is usually stated, if obliquely, in the project objectives or the evaluation criteria. If the criteria weight collaboration, innovation and programme above price, the client has told you what they are buying — and a response built around your rates will not score, however good the rates are. Our guide to how government tenders are scored covers reading the weightings; this is a case where they carry unusually literal meaning.

ECI: what the two stages actually are

Early contractor involvement is the model an SME is most likely to meet first, and the one most often misunderstood. It is a two-stage arrangement. You are selected on capability and approach before there is a price, you work with the client and their designers through Stage 1, and Stage 2 is the construction — which only proceeds if the parties agree on terms at the end of Stage 1.

	Stage 1	Stage 2
What you do	Buildability review, constructability input, methodology development, programme, risk workshops, early works and investigation, target price development	Build the job
How you are paid	A fee — lump sum, schedule of rates or cost-reimbursable, depending on the arrangement	Under the agreed Stage 2 contract
What the client gets	A design that can be built, a programme they believe, and a price developed with visibility	Delivery with a contractor already up the learning curve
Your exposure	The fee is usually modest; the real cost is senior people	Normal construction risk under the agreed form

The critical structural feature: Stage 2 is not guaranteed. Almost every ECI agreement gives the client an exit — if the target price is unacceptable, if funding does not materialise, if the parties cannot agree, the client can take the developed design to open tender or shelve the project. Some arrangements make that exit expensive for the client, most do not. Treat Stage 2 as likely rather than certain, and price Stage 1 on the assumption it might be all you get.

Getting paid for Stage 1 — and what it is worth

Stage 1 fees are usually small relative to the construction value, and contractors routinely under-price them by looking at the wrong resource. The work is not done by estimators between other bids. It is done by your most experienced construction people — the ones who know how the job will actually be built — sitting in workshops, reviewing drawings and arguing with designers. Those are the people whose time you cannot replace, and they are being taken off delivery.

Before agreeing a Stage 1 fee, work out four things.

• Named people and days. How many days of your project manager, your senior engineer, your estimator and your plant manager. Build the fee from that, the same way you would build any labour-based price — the approach in our guide to preparing civil works cost estimates applies here even though there is no bill of quantities.
• Third-party costs. Geotechnical investigation, survey, temporary works design, specialist advice. Establish whether these sit inside your fee or are reimbursable, and whether the client procures them directly.
• Early works. Many ECIs include physical early works — investigation, service proving, site establishment, enabling works. These are usually priced separately and are real revenue.
• What happens if Stage 2 does not proceed. Some agreements pay an additional amount if the client walks. Ask; it is negotiable more often than contractors assume.

Then ask the harder question: who owns what you produce in Stage 1? If your methodology, temporary works concepts and programme logic become the client’s property and can be issued to the market when Stage 2 does not proceed, you have funded a competitor’s tender. Intellectual property provisions in ECI agreements vary widely and are frequently one-sided. This is one of the few places where paying for a lawyer’s hour before signing is unambiguously worth it.

The conversion point, where most of the risk sits

The moment Stage 1 becomes Stage 2 is where ECIs succeed or turn sour, and it is worth understanding the dynamic before you are inside it.

By the end of Stage 1 you have spent months on the job. Your people know it. You have almost certainly told the client things that improved the design. You want Stage 2, and the client knows you want it. Your negotiating position at the conversion point is weaker than it was at selection, not stronger — which is the opposite of what most contractors expect. The client has your knowledge either way. You have costs to recover only if the job proceeds.

Three things protect you.

• Agree the Stage 2 commercial framework at the start, not the end. The contract form, the margin, the preliminaries basis, the risk allocation and the process for agreeing the target should be settled in the ECI agreement. What remains open at conversion should be quantities and rates, not principles.
• Keep a written record of what changed and why. Every design decision made on your advice, every risk removed, every constructability change. This is the evidence that Stage 1 delivered value, and it is what you point at when the target price is challenged.
• Be genuinely willing to not proceed. A contractor who cannot walk away at conversion will accept a target they should not. Decide in advance what your walk-away looks like, using the same discipline as our guide to the go/no-go decision.

Target cost, pain-share and gain-share

Target cost is a payment mechanism, and it can be attached to almost any of the models above. The structure is consistent even when the labels change.

• A target outturn cost is agreed — the parties’ shared expectation of what the job will cost.
• Your actual costs are reimbursed as incurred, verified against records.
• A fee covering off-site overhead and profit is paid, usually fixed as a sum rather than a percentage, so that it does not grow when costs grow.
• The difference between target and actual is shared to an agreed ratio: gain-share if under, pain-share if over.
• The pain-share is normally capped, most often at the value of your fee — so the worst realistic outcome is that you work for nothing rather than that you fund the client’s project.

The mechanism has an appealing logic: nobody profits from an overrun, and everybody profits from efficiency. In practice, four things decide whether it works for you.

What to examine	Why it decides the outcome
How the target was set	A target built from your own estimate with agreed allowances is workable. A target set by the client’s cost plan, or driven down in negotiation, is a pain-share you have already entered.
What adjusts the target	Scope change, latent conditions, client delay and specified risk events should all adjust the target. If they do not, the mechanism is a fixed price wearing different clothes — see latent conditions in civil contracts.
What counts as reimbursable cost	The definition decides whether your site overhead, small tools, plant standing time and internal plant rates are recovered. Ambiguity here is always resolved against the contractor during an audit.
Whether the cap is real	A pain-share capped at fee is a bounded downside. A cap that excludes defects, LDs or specific risk categories is much less protective than it appears.

The single most important number is your internal plant rate. Under reimbursable cost, your own plant is usually charged at an agreed schedule rather than a market hire rate. If that schedule is set too low, every hour your own machines work erodes margin you would have made on a lump sum. Build the schedule from real ownership and operating cost — the discipline in our guides to plant and equipment finance and job costing and cost control — and be prepared to justify it line by line.

Open book: what you are actually agreeing to show

Every collaborative model involves some degree of open book, and contractors agree to it without establishing what it means. It is not one thing. It runs from you show us the build-up of the price to we audit your accounting system, and the difference is substantial.

Establish, in writing, four boundaries before you agree.

• Scope. Project costs only, or company overhead and profitability? The first is normal; the second is not, and should be resisted.
• Depth. Summary cost reports, transaction-level ledgers, or supporting invoices and timesheets? Each level costs you administration time that should be priced.
• Audit rights. Who audits, how often, with what notice, and for how long after completion. Open-ended audit rights extending years past final claim are common and worth negotiating down.
• Confidentiality. Your subcontractor and supplier rates are commercially sensitive. A client with visibility of your quarry pricing across a framework has information you did not intend to sell — relevant to how you structure materials supply agreements.

There is a practical prerequisite that stops many SMEs before the commercial question arises. Open book requires accounting that can produce clean, job-coded, defensible cost at any moment. If your cost coding is approximate, if plant time is reconstructed at month end, if labour is allocated by memory, an open-book contract will expose that and every disputed allocation will be resolved against you. Fix the system before you take the work, not during it.

Alliances, and why SMEs are rarely in them

An alliance is the far end of the spectrum. The owner and the commercial participants form a single integrated team, agree a target, share pain and gain, and — the defining feature — contract on a no-blame or limited-recourse basis, giving up most rights to sue each other. Decisions are made unanimously by a leadership team. There are no variations in the ordinary sense, because there is no fixed scope to vary from.

Alliances are used on large, complex, high-uncertainty programmes, and the reason SMEs are rarely commercial participants is structural rather than snobbery.

• The selection process is long and expensive. Multi-stage, with workshops, interviews and behavioural assessment running over months, and bid costs that only a large business can absorb.
• Full-time secondment is required. Alliance teams are co-located. Committing named senior staff full-time for years is not available to a business where those people are also running the rest of the company.
• Balance sheet. Pain-share exposure, even capped, must be credible against the project’s scale — the assessment covered in our guide to demonstrating financial capacity.
• Insurance is arranged at project level in ways that assume corporate participants.

That does not put alliance work out of reach — it changes the door you use. Alliances subcontract heavily, and their subcontract packages are often larger and longer than equivalent agency work. The alliance is also unusually motivated to have subcontractors who behave collaboratively, because the alliance carries the consequences of a subcontractor’s disputes. Approach it the way you would approach any head contractor, using the routes in our guide to subcontracting to Tier 1 civil contractors — but expect the alliance to ask about your systems and your behaviour in more depth than a builder would.

Managing contractor and incentivised construct-only

Two models sit in the middle of the spectrum and are more accessible than their names suggest.

Managing contractor arrangements have the contractor coordinating and letting trade packages on the client’s behalf, usually for a management fee, with the client carrying most of the cost risk. For an SME the relevant version is not the head role but the package role: the managing contractor is buying civil packages, and buying them on a basis closer to a negotiated rate than an open tender. The relationship is with the managing contractor’s package manager, and it behaves like panel work — see winning work off panels and standing offers.

Incentivised construct-only is the most SME-accessible collaborative mechanism, and it is easy to miss because it looks like an ordinary tender. The job is designed, priced and let conventionally, with an added mechanism: savings against a defined baseline, or achievement of specified outcomes, produce a shared benefit. Sometimes it is a value-engineering share, sometimes a programme incentive, sometimes a set of key result areas covering safety, environment or community.

Two cautions. Incentives tied to subjective assessment by the superintendent are worth little at bid stage and should not be priced into your margin. And incentives tied to outcomes you do not control — traffic impact, third-party approvals, stakeholder satisfaction — transfer risk while looking like an opportunity. Price the base job so it stands alone, and treat any incentive as upside.

What the tender looks like when price is not the question

A collaborative tender asks for things a lump sum tender does not, and contractors who submit their standard response score poorly. The differences are consistent enough to plan for.

What is asked	What they are testing	What fails
Commercial framework rather than a price — fee percentages, plant rate schedules, labour on-costs, overhead basis	Whether your cost structure is transparent and defensible	Round numbers with no build-up; rates you cannot justify under audit
Approach to collaboration	Whether you have actually done it, or are describing an aspiration	Generic statements about partnership and open communication
Named individuals with availability	Whether the people in the response are the people who will turn up	An org chart of everyone in the business — see key personnel CVs and org charts
Risk and opportunity workshop input	Whether you can identify risk without immediately pricing it away	A qualification list; a register copied from the last job — see the tender risk register
Evidence of behaviour under pressure	How you behave when a job goes wrong	Referees who can only confirm the job was completed — see referees and past project experience

The hardest of these for most contractors is the honest failure story. Collaborative selection processes very often ask what went wrong on a recent job and what you did about it. The answer that scores is specific, admits the problem, and describes the response and the change that followed. The answer that fails is a claim that nothing has gone wrong — which every evaluator reads as either inexperience or evasion.

Workshops and interviews: the selection you cannot write your way through

Collaborative procurement selects on people, and the selection is usually done in a room rather than on paper. Expect a facilitated workshop, a technical interview, or both — sometimes with a working session on a real problem from the project, observed by assessors watching how your team behaves rather than what it concludes.

Four practical points, beyond the general preparation in our guide to tender interviews and presentations.

• Send the delivery team, not the bid team. Assessors are meeting the people who will be in the room for the next two years. A polished bid manager who then disappears is a negative signal.
• Let the site people talk. The most valuable contribution an SME brings to an ECI is the person who has built the thing before. Rehearsing them into corporate language removes the advantage.
• Disagree well. In a workshop, an assessor learns more from how you handle a challenge than from your answer. Conceding instantly reads as compliance without judgement; digging in reads as difficulty. Test the point, give ground on evidence.
• Do not solve the problem in silence. Behavioural assessment cannot score reasoning it cannot hear.

The four risks an SME carries into these models

Collaborative contracting is sold on its upside. The downside is real and specific.

• Senior resource commitment with no guaranteed return. Stage 1 and long selection processes consume the people who make the rest of the business work. A contractor who commits a project manager to an ECI for six months has taken capacity out of delivery, and if Stage 2 does not proceed, that capacity is simply gone.
• Cash flow shaped differently. Reimbursable cost with monthly verification can pay faster than a progress claim regime, or considerably slower when the verification process is contested. Model it before you commit — the approach in our guide to cash flow in civil construction contracts applies, with the added variable that cost you cannot substantiate is cost you do not get paid for that month.
• Administrative load. Open book, target adjustment, gain-share reconciliation and audit are administration that a lump sum job does not carry. Price it as a real cost and resource it, or it will be absorbed by the project manager at the expense of running the job.
• Concentration. A long collaborative contract with one client is an excellent revenue base and a serious dependency. The framework in our guide to scaling a civil contracting business applies here with particular force.

Where SMEs genuinely win collaborative work

Setting aside the alliance end of the spectrum, there are four places where a mid-sized civil contractor realistically wins this work.

• Water authority and utility frameworks. Long-duration programme delivery, often with target cost or incentivised mechanisms, and often with SME-scale packages inside them. Covered in our guides to water authority panels and water and sewer pipeline tenders.
• Council ECI on constrained sites. Councils increasingly use early involvement on jobs where the constraint is community, traffic or an operating asset rather than engineering complexity. The values are within SME range and the competition is local.
• Renewal and maintenance programmes. Where scope emerges over the term, the commercial model is necessarily collaborative even when it is not labelled that way — see term maintenance contracts.
• Subcontract packages inside collaborative head contracts. The largest volume by far, and the one requiring the least change to how you operate — though the head contract’s mechanisms will shape your package.

One point about that last route, because it catches contractors out. If the head contract is target cost and your subcontract is lump sum, you are the only party in the chain carrying fixed-price risk. That is not necessarily bad — it is often exactly what you want — but understand that the head contractor’s incentives around your variations, your delays and your claims are shaped by whether your cost lands inside or outside their target. Read the flow-down provisions with that in mind, and read our guide to security of payment before assuming the payment regime is what the head contract says it is.

Building the credentials before the opportunity

Collaborative work is not won by responding to an advertisement. By the time a collaborative tender is issued, the client usually knows which contractors they hope will bid. Positioning is a twelve-month activity, not a two-week one.

• Get the cost system audit-ready. Job-coded, current, reconcilable. This is the single largest barrier and the one most within your control.
• Build a defensible internal plant rate schedule. Ownership cost, operating cost, utilisation assumption, written down. You will need it, and building it under time pressure guarantees it is wrong.
• Capture what you contributed, not just what you built. Every time you propose a method change that saved the client money or time, write it down with the outcome. This becomes your collaborative track record, and it is the thing SMEs have but cannot evidence.
• Maintain relationships with client engineers and designers, not just procurement. Collaborative models are chosen by technical people.
• Have the certifications in place. Quality, environment and safety certification is a common precondition — see the prequalification trifecta and the relevant scheme in our guide to prequalification by state and territory.

The underlying point is simple. These models buy a contractor whose numbers can be examined and whose people can be worked alongside. Everything above is a way of being able to prove both before you are asked.

Checklist

• Have you identified which delivery model this tender actually uses, separately from the contract form?
• Do you know why the client chose it, and does your response address that reason?
• For an ECI, is the Stage 1 fee built from named people and days rather than a percentage?
• Have you established who owns the intellectual property produced in Stage 1?
• Is the Stage 2 commercial framework agreed at the start, leaving only quantities and rates open at conversion?
• Do you know what happens, and what you are paid, if Stage 2 does not proceed?
• Under a target cost, do you know how the target was built and what adjusts it?
• Is the pain-share capped, and does the cap have exclusions?
• Is your internal plant rate schedule built from real ownership and operating cost, and can you defend it line by line?
• Have you defined the scope, depth, audit rights and confidentiality boundaries of open book?
• Can your accounting system produce clean job-coded cost on demand, today?
• Have you priced the administrative load of open-book reporting as a real cost?
• Are the people named in the response the people who will attend the workshops and run the job?
• Do you have a specific, honest account of something that went wrong and what you changed?
• If you are subcontracting into a collaborative head contract, do you know how the head’s mechanism affects your variations and payment?
• Have you modelled the cash flow of reimbursable payment with contested verification?
• Is your walk-away point defined before you enter the process?

Sources and further reading

This guide is general information for Australian civil construction businesses and is not legal, contractual, financial or accounting advice. Early contractor involvement agreements, target cost mechanisms, alliance agreements and incentive schedules are bespoke documents that differ substantially between clients and projects; nothing here describes the terms of any particular arrangement. Whether a pain-share cap is effective, what an audit right permits, who owns intellectual property developed during a first stage, and whether an incentive is enforceable are all matters for advice on the specific documents. Open-book and reimbursable-cost arrangements have tax and accounting consequences that should be discussed with your accountant. Take advice from a construction lawyer on any collaborative or two-stage agreement before signing, and read the executed contract in full rather than relying on the model’s label.

• Collaborative and relationship-based delivery models as used by Australian public infrastructure clients — early contractor involvement, managing contractor, target cost with pain-share gain-share, and project alliancing — described in §01 and §03 to §09. These are procurement practices rather than a single regulated framework, and the terminology varies between jurisdictions and agencies; the descriptions here reflect the common structure rather than any one client’s model. Individual state transport, water and infrastructure agencies publish their own delivery model guidance, and the applicable version is the one referenced in the tender documents.
• Standard-form construction contracts, which remain the legal basis for the construction stage of every model described here. The forms, their risk allocation and the effect of amendments are sourced in full in our guides to AS 4000 and AS 2124 and contract forms beyond construct-only, and the pricing models are covered in schedule of rates versus lump sum versus cost-plus.
• The commercial and cash flow observations in §05, §06 and §12 are drawn from the mechanics of reimbursable-cost contracting rather than from published survey data, and are presented as reasoning rather than measurement. The internal plant rate discussion assumes ownership and operating cost is built as set out in our guides to plant and equipment finance and job costing and cost control.
• Related TenderBuilt guides carrying the primary-source detail referenced above: how government tenders are scored, the tender risk register, latent conditions, demonstrating financial capacity, key personnel CVs and org charts, tender interviews and presentations, subcontracting to Tier 1 contractors, winning work off panels and standing offers, term maintenance contracts, water authority panels, cash flow in civil construction contracts and security of payment in Australia.