Before the Evidence Arrives: Board Discipline for Joint Innovation Pilots

Joint innovation pilots have become a familiar route for UK organisations testing new services, technologies and delivery models. A local authority may trial a digital referral pathway with a charity; an NHS trust may test a remote monitoring tool with a technology supplier; a private firm may explore a new logistics model with a university or an arm's-length body. The appeal is understandable. Pilots promise learning without full commitment, evidence before scale, and collaboration without the weight of a long-term contract.

That promise is also the governance problem. Because pilots are framed as provisional, boards often approve them lightly and oversight then rests on the enthusiasm of the project team. Ownership of intellectual property, data, costs and decision rights is deferred until something succeeds or fails. By then, the organisation may be negotiating from a weaker position, or discovering that a trial has already become business as usual without a deliberate decision.

The board's task is not to run the pilot. It is to ensure that the trial has a clear owner, a defined purpose and a pre-agreed route to a decision. A one-page pilot charter can provide that discipline without smothering experimentation.

Why pilots slip past governance

Pilots frequently sit below the threshold that triggers formal partnership governance. They may be described as discovery work, a proof of concept or a learning exercise. Budgets are often modest, timescales short and the language deliberately provisional. Each of these features is reasonable in isolation. Together, they can create a governance gap.

The gap matters because a pilot is rarely neutral. It consumes staff time, creates expectations among service users and partners, and generates data and intellectual property. If the trial succeeds, the parties must decide who owns the resulting asset, how it is licensed, who pays for transition and whether the public-sector partner can scale it within procurement rules. If it fails, they must decide who bears the sunk cost, what happens to shared data and how the ending is communicated.

In the UK public sector, these questions are sharpened by procurement and subsidy control requirements. A pilot should not become a route around competition or transparency. Boards should seek assurance that the procurement route has been considered at design stage, not after the trial has produced a promising result.

The one-page charter

A pilot charter need not be a lengthy legal document. Its purpose is to force the board and the executive to answer a small number of questions before work begins. It should be short enough to be read, specific enough to be tested, and signed off by a named board sponsor.

The charter should cover seven fields:

• Purpose and problem: what the pilot is intended to learn, and what decision the evidence will inform.
• Parties and roles: who is involved, who leads, who supplies what, and who speaks for each party.
• Decision rights: which decisions sit with the board, the executive, the project board or the partner, and which require unanimous agreement.
• Intellectual property and data: who owns foreground and background IP, what data is shared, on what lawful basis, and what happens to it at the end.
• Costs and liabilities: the budget, the cost-sharing basis, the treatment of overspend, and the limits of liability.
• Stage gates and success criteria: the points at which the board will review progress and the evidence required to continue.
• Exit and transition: the default sunset date, the route to closure, and the conditions under which the pilot may become business as usual.

This is not a substitute for legal advice or a formal contract where one is needed. It is a governance instrument that makes the board's expectations visible and gives the executive a framework for escalation.

Stage gates that force a decision

A pilot without gates tends to drift. The board should agree a small number of review points, each with a clear question. At approval, the question is whether the pilot is worth starting and whether the charter is adequate. At design, it is whether the trial is properly scoped, lawful and resourced. At midpoint, it is whether the evidence is emerging as expected and whether risks remain tolerable. At the evidence gate, it is whether the success criteria have been met. At transition, it is whether the organisation should stop, extend, procure or scale.

The board does not need to review every detail. It needs to know who is accountable for bringing the gate decision and what happens if the gate is missed. A named board sponsor, with a defined reporting line, is often the difference between a pilot that is governed and one that is merely observed.

Success criteria beyond the demonstration

Pilots are often judged by whether the technology works or the demo impresses. Boards should ask for a broader test. Does the pilot improve outcomes for service users? Is it operationally deliverable at scale? Does it create a credible commercial case? Are the risks to data, reputation and continuity acceptable? Has the partner demonstrated the capability and capacity to continue?

It is equally important to define failure criteria. A pilot that cannot fail is not a pilot. The charter should state what evidence would cause the organisation to stop, and who has authority to make that call.

Procurement, subsidy control and the public-sector partner

Where a public-sector body is involved, the board should seek assurance that the pilot has been designed with procurement and subsidy control in mind. Early market engagement, documented decision-making and a clear record of the route chosen will matter if the pilot later moves to scale. The board should not assume that a trial is automatically outside the UK's procurement regime. Legal and procurement advisers should be involved at the charter stage, not only at contract stage.

The exit route

Every pilot charter should include a default sunset date. If the parties wish to continue, they should return to the board with a transition plan. That plan should address ownership and licensing of IP, return or deletion of data, treatment of residual costs, continuity of service, communications with users and staff, and the procurement route for any ongoing arrangement.

The board should ask two questions before approval. If the pilot works, what have we already agreed about ownership and scale? If it fails, who pays and how do we close it well? If neither question can be answered, the charter is not yet ready.

A small instrument with a large effect

Joint innovation is valuable, but it is not exempt from governance. A one-page pilot charter gives the board a proportionate way to approve experimentation while retaining control of the decisions that matter. It protects the relationship as well as the organisation, because it reduces the chance that success or failure will be settled in hindsight, under pressure, with the terms unclear.

For boards, the discipline is simple: before the evidence arrives, agree what it will be used for, who owns it and how the trial will end. That is not bureaucracy. It is the condition for confident collaboration.