How to Structure Client Delivery in a Professional Services Business

Consistency by person does not scale

In a small professional services firm, delivery is consistent because the same two or three people are involved in everything. They know what good looks like, they catch problems, and clients get a reliable experience.

The failure mode is not dramatic. As the firm grows, those people are spread across more engagements, and quality becomes a function of who happened to be available. Nothing breaks visibly. The variance just widens, and the firm finds out through a client rather than through a system.

The instinct is to hire more senior people. That works and it is expensive. The cheaper first move is to put into the structure whatever those people were supplying from memory.

The engagement is the unit

A professional services firm does not sell tasks and does not really sell projects. It sells engagements: a commitment to a client, with a scope, a value, a duration and a team.

Making the engagement the primary unit has a specific consequence. Every question worth asking has somewhere to be answered:

• What did we commit to, and at what price?
• Who is delivering it?
• What has been produced so far, and what has been signed off?
• What has it cost, and what is it going to cost?
• What have we billed, and what has been paid?

If those five answers live in five systems, nobody asks them together, which means nobody notices when the answers stop agreeing.

A client can hold several engagements at once, and they should be separate records rather than one large one. A firm running a discovery, an implementation and a support retainer for the same client is running three commercial commitments with three different economics.

Give engagements a shape

Most firms deliver a small number of engagement shapes repeatedly, even when it does not feel that way. A discovery. An assessment. An implementation. A retained advisory arrangement.

For each shape, decide once what it normally contains: which deliverables, in what order, with what typically going to the client at each stage. This is not a rigid methodology, it is a starting point that stops every engagement being designed from nothing by whoever is free.

The same applies one level down, to the deliverables themselves. For each kind of deliverable your firm produces regularly, decide what information the person producing it must record. A workshop needs the date, the attendees and the outputs. An analysis needs its inputs and assumptions. Deciding once is what makes the record consistent regardless of who produced it.

Separate producing from reviewing

Every deliverable needs one person producing it and one person deciding it is good enough. In professional services this is often the only thing standing between the firm and a reputational problem, and it is frequently informal.

Formalising it costs very little. Name a reviewer per engagement so the default is set, allow it to be overridden per deliverable, and record the decision when it is made: who, when, what they decided, what they said.

The value shows up in three places. Junior people know who to hand work to. Clients get a consistent standard rather than one that depends on who was available. And when something does go wrong, the firm can see where, rather than having a conversation about impressions.

Worth being strict about one thing: the producer and the reviewer must be different people. A sign-off step somebody performs on their own work is a habit, not a control.

Attach cost to work, not to time

Firms that bill by time naturally track cost by time, and for them the two line up. Plenty of professional services firms do not: fixed-fee engagements, retained arrangements, output-based subcontracting, associates on monthly retainers.

For those, insisting that cost equals hours times rate means asking people to record hours nobody is billing, which produces resented data of poor quality. The alternative is to let cost attach to the work in whatever way is true for the person doing it:

• An associate on a monthly retainer costs a monthly amount, regardless of which engagement they touched.
• A specialist brought in for one deliverable costs whatever was agreed for that deliverable.
• A subcontractor paid by the hour costs hours times rate.

All three can be true inside the same engagement. What matters is that each piece of work knows how it is costed, and that the moment cost becomes real is defined. Approval is usually the right moment, because it is when somebody with authority accepted the work.

One honest limitation. Period-based costs, such as a monthly retainer for somebody working across six engagements, cannot be allocated to individual deliverables without an assumption. Any allocation formula you pick will be defensible and slightly wrong. It is usually better to report those costs at the level they were incurred than to spread them and produce a per-engagement figure that looks precise and is not.

Keep operational and commercial apart

Scope changes on a professional services engagement almost always start operationally: somebody discovers the data is worse than expected, or the client asks for one more session.

Two separate things should now happen, and firms routinely collapse them:

• The operational record. Extra work exists and will cost something. Whoever discovered it should be able to record it immediately, without permission.
• The commercial decision. Whether the client pays more. That requires a conversation and a deliberate adjustment to the agreement.

If the first automatically triggers the second, you bill clients for things nobody agreed. If recording extra work requires the commercial conversation to happen first, people stop recording it, and the firm loses visibility of exactly the thing that erodes its margins.

When the agreement is adjusted, keep the original figure. A firm that can see an engagement was renegotiated from 40,000 to 55,000 knows something about its scoping that a firm seeing only 55,000 does not.

Client visibility as a deliverable

Professional services clients are usually senior, busy, and reporting internally on the work they have commissioned. What they need is not access to your delivery process; it is the ability to answer their own internal question without emailing you.

Which usually means four things: what is in progress, what has been delivered, where the documents are, and what has been invoiced. It rarely means seeing which associate produced what or how many internal review cycles a deliverable took.

Deciding to show less is not a lack of transparency. Internal review is a process for producing quality, not a fact about the engagement, and exposing it invites clients into conversations about how you work rather than what you delivered.

A note on utilisation

Nothing in this guide covers utilisation, capacity planning or resource forecasting. That is deliberate.

Those are a genuinely different class of problem, and they matter enormously to firms above a certain size, where the central question is whether you have the right people available in eight weeks. Solving that properly needs dedicated tooling, and firms that need it know they do.

For a firm below that threshold, the more valuable question is not whether people are busy. It is whether the work they are busy on is making money, which is what everything above is about. Plenty of firms buy sophisticated capacity planning and still cannot say which of their engagements is unprofitable.

Where to start

In order, because each makes the next easier:

1. Make the engagement a real record carrying the agreed value, currency and team. Not a folder, a record.
2. Name a reviewer per engagement. One decision that changes the default for every deliverable inside it.
3. Record the review decision when it is made, rather than in a message.
4. Define how each kind of person is costed and when cost becomes real.
5. Separate additional work from agreed scope at the point work is created.
6. Give clients a view that answers their four questions, last, once the rest works.

A firm that does the first three has made delivery consistent independently of who is running it. That is the part that was holding growth back.