Escalation by Design: Giving UK Clients a Route That Is Not a Complaint

There is a line in most UK service contracts that reads roughly as follows: the parties shall use their reasonable endeavours to resolve any dispute through the agreed account management structure. It is a sentence drafted to satisfy a lawyer and destined to be forgotten by everybody else. It describes an intention, not a route, and it names no one.

When something goes wrong at month seven, the client does not consult the contract. They telephone the person they already know, or they say nothing and let the dissatisfaction accumulate. Both responses are rational. Neither is good governance.

A clause is not a route

The distinction matters more than it sounds. A clause allocates risk after the fact. A route is an operating mechanism that exists before anything has gone wrong, with named people, defined thresholds and an agreed rhythm.

Organisations that get this right treat escalation as part of service design rather than as a legal backstop. They accept a plain truth: in any relationship of substance, something will eventually go wrong. The question is not whether the disagreement occurs but how many weeks pass before the right two people are in a room together.

Why small issues become formal disputes

The pathway from minor irritation to formal dispute is remarkably consistent. A deliverable arrives slightly late. The client mentions it to the engagement lead, who is apologetic but busy. The next one is later still. By the fourth month the client has concluded that the supplier does not care, and the supplier has concluded that the client is difficult to please. Neither has said any of this out loud.

By the time the issue reaches a partner or a procurement lead, it has usually stopped being about the deliverable. It is about the story each side has told itself. That story is expensive to unpick and it rarely survives a well-run conversation in month two.

Crucially, the failure here is structural rather than personal. Most clients do not want to complain. Complaining feels like an admission that they chose badly, and it invites a defensiveness that makes the next six months harder. So they absorb the problem until they cannot.

Name people, not departments

A workable escalation route begins with names. Not roles in the abstract — the service desk, the account team, the partner group — but two or three individuals with telephone numbers and an explicit licence to be contacted.

A structure that works in practice has three tiers and no more. The first is the day-to-day relationship: the person the client already deals with, who should be reachable and empowered to fix ordinary problems without seeking permission. The second is a named senior contact on each side, used when the first tier has not resolved something within an agreed period — five working days is a common and sensible threshold. The third is a single executive sponsor on each side, engaged sparingly, whose involvement signals that the matter is now serious.

Three tiers is enough. A four- or five-tier structure looks thorough on paper and collapses in use, because nobody can remember which tier they are on.

Thresholds that can be applied on a Friday afternoon

Escalation triggers should be written so that a busy person can apply them without interpretation. Elapsed time is the most reliable: more than five working days without resolution at tier one. So is a missed commitment that affects a downstream date the client has already published internally. So is any request to change scope by more than an agreed percentage.

Vaguer triggers — a material deterioration in service quality, a failure of confidence — are common in contracts and almost never invoked, precisely because they require somebody to make a judgement call about a relationship they depend on.

The best test of a threshold is whether the client could apply it while tired, irritated and short of time. If it requires a meeting to decide whether the threshold has been met, it is not a threshold.

The scheduled difficult conversation

Even well-designed routes go unused, because human beings avoid friction. The most effective correction is not more process but a diary entry.

A short, standing review — thirty minutes, monthly or quarterly depending on the intensity of the service — with one question on the agenda ahead of the usual performance items: what is not working as well as it should? The framing matters. Asked inside a scheduled conversation, the question invites candour. Asked in an email after a missed deadline, it invites defensiveness.

These checkpoints should be minuted lightly, with actions and owners, and they should be attended by the tier-two contacts rather than delegated. The presence of senior people signals that the conversation is safe to have.

Escalation is not failure

Many UK organisations still treat an escalation as evidence that the relationship is in trouble. That instinct is costly. A client who raises a concern at the right level, early, is doing the supplier a favour. The alternative is not harmony; it is silence followed by a competitive tender.

Suppliers should therefore say so explicitly, in writing and at the start of the engagement: we would rather hear it at tier one in week three than at tier three in month nine. Clients, for their part, should treat the existence of a used escalation route as a sign of a functioning relationship rather than a damaged one.

Proportionality and the right to be wrong

A sound route also protects the supplier. Escalation should not be a mechanism by which a client can override professional judgement on a matter within the supplier's remit. The rights and obligations on each side should be symmetrical: if the client may escalate a missed deadline, the supplier may escalate an instruction that it believes is imprudent or outside the agreed scope.

This symmetry is what separates governance from grievance. It also tends to improve the quality of the conversation, because neither party arrives at the table as the accused.

A short checklist for 2026 contracts

When the next service agreement is drafted, four questions are worth asking before signature. Are the escalation contacts named individuals, with direct contact details? Are the triggers expressed in elapsed time or observable events rather than judgement? Is there a standing review in the calendar for the first twelve months, with the difficult question on the agenda? And does the supplier have an equivalent right to escalate?

Four questions, fifteen minutes of drafting, and a materially lower chance that a small operational problem becomes a formal dispute. For most mid-market organisations, that is a better return than another page of warranties.