Ask Before You Appoint: Six Questions That Turn Good Intentions Into Sound Governance

Procurement timelines in UK organisations have a habit of compressing at the worst possible moment. A business case is approved late in the quarter, a supplier list is drawn up in a fortnight, and the pressure to appoint quickly begins to outweigh the patience needed to define what is actually being bought. None of that is unusual. What is unusual is how rarely the resulting ambiguity is priced.

At DGA Ltd we have observed that the most costly problems in a professional services engagement rarely originate in the work itself. They begin earlier, in the gap between what a client thought they had commissioned and what a supplier believed they had agreed to deliver. That gap is seldom caused by bad faith. It is usually caused by six questions nobody asked.

The audit below is not a procurement formality. It is a risk-control instrument, and it takes less than an hour to run.

One: What is in scope, and what are we deliberately leaving out?

Every credible engagement produces a scope statement. Far fewer produce an exclusions schedule. The distinction matters, because scope creep rarely arrives as an obvious request. It arrives as a reasonable-sounding favour, a small addition, a colleague's question that looks like ten minutes of work and turns into three weeks.

Ask for both documents before appointment. A written list of what sits outside the agreement, together with how a change would be priced and approved, is not a sign of a difficult supplier. It is a sign of one that has done this before. In construction-adjacent work the discipline is familiar from NEC-style contracts; in consultancy and transformation programmes it is frequently missing altogether. It should not be.

Two: Who decides, and by when?

Uncertainty about decision rights is the most common cause of a stalled programme. Two capable teams can spend months waiting for each other to approve a design, a specification or a hire, each convinced the other held the pen.

Before work begins, agree an approvals matrix written in plain English. Who can sign off workstream-level decisions? Who must be consulted but not obeyed, as distinct from who must consent? What value threshold pushes a matter to an executive or a board committee? What is the agreed turnaround for each decision, and what happens if that window passes unused?

The output need not be elaborate. A single page, naming roles rather than job titles where possible, is usually sufficient, provided it is circulated and accepted by everyone who will actually be involved.

Three: How often will we hear from you, and in what form?

Reporting cadence is where good intentions most often dissolve. Regular updates is not a cadence.

Ask for specifics: a written note every week, one page, issued by a named day, covering progress against plan, decisions required, risks and spend to date. Ask who receives it, who reads it before it is sent, and whether verbal updates will supplement rather than replace the written record.

Written updates do something meetings cannot. They create a searchable history, they force brevity, and they make drift visible early. A client who can scroll back through twelve weeks of notes holds a far stronger position than one relying on recollection of a call in February.

Four: What happens when something goes wrong?

Every engagement encounters difficulty. The relevant question is not whether, but how quickly it surfaces and through which route.

Ask for a named escalation contact, the response time they commit to, and what constitutes a trigger for escalation. Ask, too, whether the route bypasses the day-to-day team or runs through it. Both models can work, but they must be agreed rather than improvised at the point of strain.

A useful test is to ask the supplier to describe the last two occasions on which they escalated to a client, and what followed. The answer reveals more about culture than any capability statement.

Five: How will we know it worked?

Success measures agreed after delivery is complete are not measures; they are narratives. Establish the baseline before commencement. What is the current figure, the current cycle time, the current error rate, the current cost per case? Who will measure it, using which system, at what intervals?

Where the benefit is qualitative, name the evidence that will be accepted. A board paper, an audit finding, a user research panel, an independent review: whichever is chosen, choose it in advance.

It also helps to separate outputs from outcomes. A supplier can control deliverables; they can only influence results. Being clear about which is which protects both parties from an uncomfortable conversation later.

Six: How does this end?

Exit terms are usually negotiated at signature and rarely revisited, which is precisely why they deserve attention while leverage still exists.

Ask about notice periods, knowledge transfer, documentation standards, the return or destruction of data under UK GDPR, ownership of intellectual property, and the handover plan for the final weeks. Ask what an orderly transition looks like and who bears the cost. Ask whether the people who delivered will still be available during handover.

An engagement with a clear ending is easier to start. Clients who know how to leave tend to be more confident about staying.

Why the hour is worth spending

The six questions above take less than an hour in a well-prepared meeting. The ambiguities they prevent typically cost considerably more, in rework, in governance friction, and in the slow erosion of trust that follows a misaligned expectation.

None of this requires an adversarial posture. Most suppliers welcome a client who asks precise questions, because recycled ambiguity is as expensive for them as it is for you. What it requires is a decision to treat clarity as part of the brief rather than a hoped-for byproduct.

At DGA Ltd we work with UK organisations that prefer to settle these matters early. The conversation is not complicated, but it is worth having before the contract rather than after it.