What Client Management Actually Means in Professional Services

Client management, in a firm that sells its people rather than a product, is the practice of running an engagement from the decision to accept it through to the decision that it is finished, in a way that holds up when somebody asks a question about it eighteen months later. That last clause is the whole difficulty. A firm can deliver excellent work and still be unable to show what it agreed, what changed, who approved the change, what was handed over, or why a conclusion was reached. When a client disputes an invoice, when a regulator asks how independence was assessed, or when a partner leaves and takes the context with them, the record is the only thing left.

This is why client management is a different discipline from customer relationship management, and why treating them as one produces software that is good at neither. A CRM is built around the pursuit: the account, the pipeline, the contact, the next conversation. It is optimised for the period before a signature. Client management begins where that ends, and its objects are not leads and opportunities but engagements, scopes, deliverables, changes, risks, and the evidence behind each of them.

The five things a client management system has to hold

Strip away the feature lists and there are five records that a professional services firm cannot operate without. Any tool that claims this category should be judged on whether it holds all five and connects them, or whether it holds one and gestures at the rest.

• The engagement itself: who the client is, what was agreed, the contract value, the dates, and the current health of the work. This is the spine everything else hangs from.
• The acceptance decision: whether the firm may take this work on at all, and the checks that answer that question. In regulated work this is not optional and not a formality.
• The scope and its baseline: what was promised, when, and what the plan looked like at the moment it was agreed, so that later movement is visible as movement rather than absorbed silently.
• The change record: every request to alter scope, schedule or fee, its impact, who approved it, and what it did to the baseline and the commercial position.
• The evidence file: deliverables and their sign-offs, meetings and their minutes, decisions and who took them, and the working papers behind the conclusions.

Why the pieces are only useful together

A status report that contradicts the risk register is worse than no status report, because it teaches the reader that the reports are decoration. This is the most common failure in firms that run each of these five records in a separate place: the register lives in a spreadsheet, the status report is written in a document from memory, the change requests are in email, and the deliverable sign-offs are attachments somebody saved to a shared drive. Each artefact is individually defensible and collectively they disagree.

The practical test is whether a number can be traced. If a status report says the engagement is amber on budget, somebody should be able to click through to the reason: the approved changes that moved the fee, the expenses claimed against it, the effort recorded, and the milestone that has not been invoiced. If that trace does not exist, the amber is an opinion, and opinions are what get argued with at a steering committee rather than acted on.

The same logic applies to the risk register and the report that quotes it. When the key risks in a weekly report are selected from the live register rather than retyped, the two cannot drift. When they are retyped, they drift within a month, and the drift is invisible until somebody notices that a risk closed in March is still being reported in June.

What changes when the client can see it

A client portal changes the incentives inside a firm more than it changes the experience outside it. Once a client can see the information requests that are outstanding, the deliverables awaiting their sign-off, and the change requests waiting on their decision, the internal record has to be accurate, because the client is reading it. Firms that adopt a portal usually discover that their internal hygiene was worse than they believed, and that discovery is the main benefit in the first quarter.

It also moves an argument that used to happen at the end of an engagement to the middle of it. A client who signs off a deliverable at the moment it is delivered cannot reasonably reopen it at closure, and a client who is shown a change request with its cost and schedule impact at the moment it is raised cannot reasonably claim the work was assumed to be included. Neither of those is about software. Both of them depend on a record the client actually saw, on a date, with what they were shown preserved.

How to tell whether you have a client management problem

Any two of these together are enough to justify changing how the work is run. All six is the normal state of a firm that has grown past the point where one person could hold an engagement in their head, and has not yet replaced that person with a system.

• Somebody has to ask two different people to answer the question of what was agreed on a specific point of scope.
• The weekly status report is written from memory rather than assembled from the registers, and takes more than twenty minutes.
• A change was delivered and never priced, or priced and never approved, and nobody can say which.
• A deliverable was accepted verbally and there is no record of who accepted it or when.
• The engagement margin is known at the end of the month and not during it.
• A partner joining an engagement mid-flight spends their first week reconstructing context from email.

Choosing where to start

The instinct is to start with the plan, because planning software is familiar and a Gantt chart looks like progress. In practice the highest return is at the two ends. Acceptance, because a badly accepted engagement is a problem that compounds for its whole life and cannot be fixed later. And closure, because a firm that never formally closes an engagement never learns anything from it, and repeats the same mispricing on the next one.

The middle takes care of itself once those two ends are firm, provided the registers are shared rather than duplicated. A firm that gets acceptance and closure right and keeps one register per thing is running better client management than a firm with an elaborate plan and four copies of the truth.