Teaming Up Without the Tangle: The Compliance Side of Subcontractor Alliances

Ask a group of small builders what is holding them back in 2026 and you will usually hear the same answer: the work they want is one size up from the work they get. A three-van outfit in the Midlands can comfortably deliver a £250k extension package. The £1.4m school refurbishment down the road needs a principal contractor with a construction phase plan, a full insurance schedule and enough balance sheet to satisfy a pre-qualification questionnaire. So the job goes to a firm with forty staff, and the small firm stays small.

More contractors are now sidestepping that trap by joining forces. Not merging, not selling out — just forming loose alliances with other builders and specialist trades so they can bid as a single team. Done well it works brilliantly. Done badly, it is how you end up liable for somebody else's shoddy fire-stopping and paying their labourers out of your own bank account.

This piece is about the second part. Winning the work is the bit everyone gets excited about. The compliance is where the grown-up work happens.

Why the mood has shifted towards teaming up

Three things have changed. Main contractors are carrying fewer, larger packages because it is cheaper to manage six subcontractors than sixteen. Public sector buyers, nudged along by the Construction Playbook, want evidence of capability and a single point of responsibility on the contract. And the cost of bidding — estimating, design development, pre-construction hours — has climbed to the point where a small firm genuinely cannot afford to write off five unsuccessful tenders a year.

Teaming up spreads all three problems. You share the bid cost, you present a combined capability statement, and you give the client one throat to choke. The catch is that the client will also treat you as one entity when things go wrong.

First question: who is the client actually contracting with?

There are broadly three shapes, and they carry very different risk profiles.

The simplest is a lead contractor model. One firm signs the contract as principal contractor and the others sit beneath it on domestic subcontracts. Clean, familiar, and the lead party carries the risk — which is why the lead party usually wants a bigger slice of the margin.

Next is a formal joint venture, often an unincorporated JV or a special purpose company set up for one project. This is where joint and several liability bites. If your partner walks off site owing three subbies £40k, the client can come after you for the lot, then you argue about it later. That is not a reason to avoid JVs — it is a reason to know exactly who you are in bed with.

Third is the informal consortium, typically used for frameworks or a series of smaller packages. It is the least legally heavy and the most likely to go wrong quietly, because nobody writes anything down.

Whichever shape you pick, put it in writing before the tender goes in. A one-page heads of terms signed by every partner is worth more than a forty-page agreement drafted after the dispute starts.

Insurance is where teaming arrangements quietly fall apart

This is the single most common gap we hear about. A builder checks their public liability cover, sees £5m, and assumes they are fine. Then you read the policy and find the limit is £5m in the aggregate across the year, not per claim, and that contract works you are delivering on behalf of another entity are excluded unless notified.

Four things to check before you bid together:

• Public and products liability limits — per claim or aggregate? Is the limit adequate for the project value the client is asking for?
• Contractors all risks — who is insuring the works, and does the policy name all the parties? A JV that nobody told the insurer about can be a very expensive afternoon.
• Professional indemnity — if there is any design element, even design-and-build on a fit-out, somebody needs PI. Check whether it is on a claims-made basis and whether the retroactive date covers the project.
• Cross-liability — if you are insuring on behalf of the group, does the policy treat each party as separately insured?

Ask each partner's broker to confirm in writing that the arrangement is disclosed. If a broker hesitates, treat that as your answer.

Competence is not a vibe

Under CDM 2015, competence has to be demonstrated, not asserted. When you team up, the client and the principal designer will want to see that every member of the group can do what they claim.

In practice that means a folder per partner covering: relevant experience on comparable projects, accreditations that actually apply to the scope (NICEIC or equivalent for electrical, Gas Safe for gas, UKAS-accredited surveyors for asbestos), the qualifications held by supervisors, and how many people hold SMSTS or SSSTS tickets. Add your accident frequency rate, your last two years of RIDDOR reportable incidents, and evidence of how you induct and monitor people on site.

Most of this is the same evidence you already gather for a PQQ. The difference is that in a consortium you are answering for your partners as well. If one of them cannot produce a construction phase plan for a notifiable project, that is now your tender's problem too.

Sorting out the money before the first invoice

Cash is the other place alliances come unstuck, and it usually starts with something small — a CIS deduction nobody expected, or a retention held by the lead contractor that the subbies thought was being held by the client.

Get these agreed in writing up front:

• CIS status — who is registered, who has gross payment status, and who is verifying whom. A 20% deduction appearing on a payment you had not budgeted for can wipe out your margin on a small package.
• Retention — what percentage, who holds it, in what account, and when it is released. If the lead contractor holds it, you want to know what happens if they become insolvent.
• Payment terms — align them to the head contract. If the client pays the lead on 45 days, subcontractors on seven-day terms are lending the project money.
• Variations — who authorises them. In a consortium with no single decision-maker, a verbal instruction from a partner's site manager is not a variation.
• Bonds and guarantees — if a performance bond is required, who is funding it and who is exposed if it is called?

A simple teaming agreement checklist

You do not need a City law firm for a single-project alliance. You do need these points covered in writing, signed by everyone:

1. The parties, their legal status and Companies House numbers.
2. The project and the scope each member is delivering.
3. Who signs the head contract and in what capacity.
4. How the bid cost is shared and what happens if the bid fails.
5. Who holds the retention and any bond.
6. Insurance responsibilities, limits and evidence deadlines.
7. Payment flow — dates, CIS treatment, what triggers an invoice.
8. Who authorises variations and how much they can approve.
9. Health and safety responsibilities, including CDM duty holders by name.
10. Programme obligations and the consequences of a partner running late.
11. A step-out clause: notice period, how a partner exits, how their work is valued.
12. Dispute route — adjudication, mediation, or a named expert.

And the exit route

Most alliances do not fail on the first project. They fail on the third, when one partner has grown faster than the others, or is quietly undercutting the group on the same client. Build in a review point — a short conversation at project close, plus a documented right to step away from the next bid. It keeps things human, and it keeps the paperwork honest.

Teaming up is the most realistic route to bigger contracts for a small UK builder in 2026. It just needs to be the boring, documented version of teaming up.