The Dependency Test: When One Client Starts Setting the Agenda

Most professional-services firms can name their largest client. Fewer can describe, with any precision, how much of their strategy that client has quietly come to own. Concentration is usually reported as a percentage of fee income. That figure matters, but it is a crude instrument. It says little about margin dependency, cash exposure, key-person risk or the range of decisions a firm has begun to take for granted because one relationship is at stake.

For a UK board, the question is not whether a dominant client is good or bad. It is whether the dependency is visible, priced and governed. A client that provides stability can also narrow the firm's options. The board's task is to ensure the second effect is understood before it becomes structural.

Revenue is the least interesting number

A client accounting for 15 per cent of fee income may look manageable. If it contributes 40 per cent of gross margin, occupies 60 per cent of one team's available hours, or determines the firm's working-capital cycle, the headline revenue figure understates the exposure. Boards should look at the client at the level at which it actually affects the business: practice group, office, service line and cash.

Useful measures include client-level gross margin after direct costs, work in progress, debtor days, utilisation of named staff, and the proportion of discretionary investment or business development time the relationship consumes. A client that pays late but reliably is not the same as one that pays late and unpredictably. A client that generates a reference is not the same as one that generates conflicts. These distinctions should appear in the board's information, not remain in management's head.

Four forms of lock-in

Concentration is often described in commercial terms, but lock-in takes several forms.

Commercial lock-in arises when a client's fee income supports fixed costs, or when discounts and scope concessions have become habitual. The firm may believe it is protecting revenue when it is actually protecting a relationship that no longer earns an appropriate return.

Operational lock-in appears when teams, systems or processes have been shaped around one client's requirements. Bespoke reporting, dedicated account management and client-specific technology can be sensible. They become a risk when they cannot be redeployed without cost or disruption.

Reputational lock-in is less discussed. A dominant client may create conflicts, constrain the firm's ability to act for competitors, or draw the firm into public positions it would not otherwise take. The board should ask whether the relationship is enhancing the firm's standing or narrowing it.

Key-person lock-in is the most familiar and the most frequently tolerated. If one partner or director holds the relationship, the client may be loyal to the individual rather than the firm. Succession planning should be tested at client level, not only at board level.

There is also a quieter form: governance lock-in. When a client is important enough, exceptions are made. Credit terms, scope changes and escalation routes bypass normal controls because the relationship is deemed too valuable to disturb. Each exception may be defensible. Collectively, they transfer authority from the firm's governance to the relationship.

The hidden cost of saying yes

A dominant client rarely demands that a firm say yes. It simply creates conditions in which saying no feels costly. The board should make those costs explicit.

Consider a client that asks for a reduced rate on a new workstream. The immediate margin effect is visible. Less visible is the precedent, the opportunity cost of the team's time, and the message sent to other clients. Consider a client that requires a partner's presence at short notice. The cost appears in delayed business development, not in the client's P&L. Consider a client that pays on extended terms. The cost appears in the firm's cash position and its ability to invest.

A simple management discipline is a strategic exceptions log. It records client-related departures from standard terms, pricing, staffing or approval routes, together with the reason and the approver. Reviewed quarterly, the log gives the board a practical view of where one relationship is bending the firm's operating model. It is not a mechanism for blame. It is a mechanism for pricing dependency.

A boardroom exercise: mapping dependency

The exercise can be completed in a single session, provided the board has the right data and is willing to use client codes rather than names where confidentiality requires it.

Step one: rank by economic contribution. Take the top ten clients by fee income and add gross margin, debtor days, WIP, utilisation of named staff and the number of people who could credibly lead the relationship. Rank the clients again on a composite view. The purpose is not to produce a precise score but to expose clients whose importance is hidden by the revenue ranking.

Step two: test replaceability. For each client, ask what would need to happen if instructions stopped tomorrow. How many months of fee income would need to be replaced? Which roles would be affected? Are there conflicts that would prevent the firm from replacing the work quickly? Could the relationship be transferred internally without losing the client? The answers should be recorded, not debated in the abstract.

Step three: agree thresholds. Set board-approved thresholds across revenue, margin, cash and key-person exposure. The numbers should fit the firm's risk appetite; a threshold copied from another business is a distraction. As an illustration, a board might require notification when any client exceeds 20 per cent of fee income, a mitigation plan when it exceeds 30 per cent, and a formal succession or exit review when one individual holds more than a set proportion of a practice group's client relationships. The precise figures matter less than the pre-agreed response.

Step four: assign a dependency owner. The relationship partner manages the client. The board should also appoint a non-executive or committee lead to monitor concentration. This separates client care from governance oversight and makes it harder for commercial momentum to override the firm's risk appetite.

Step five: review the exceptions. Use the strategic exceptions log to test whether the agreed thresholds are being respected in practice. A board that sets a threshold and then approves exceptions without recording their cumulative effect has not set a threshold at all.

Thresholds that mean something

A threshold should trigger an action, not merely a colour on a dashboard. Amber might require a report to the board and a named mitigation plan. Red might require a pause on new commitments to the client, a review of pricing and terms, and a board-approved plan to diversify or transfer the relationship. The response should be recorded in the risk register and reviewed at fixed intervals.

The board should also test the client's own view of the relationship. A dominant client may value continuity, but it may also be uncomfortable with dependency. A structured conversation about succession, service resilience and commercial terms can strengthen the relationship rather than signal distance. The objective is not to end the relationship. It is to ensure the firm can continue to serve it on terms that remain rational.

The test that matters

Client concentration is not inherently a failure of strategy. Every firm makes choices about where to invest. The problem arises when those choices are made by default, one accepted exception at a time, without the board seeing the cumulative position.

The test is simple enough to apply in a single board session: can we describe, in concrete terms, what this client owns of our strategy, our cash, our people and our options? If the answer is unclear, the dependency is already doing work that the board has not authorised. That is the point at which a good relationship becomes a governance question.