How to scale advisory work without scaling hours

Almost every firm delivering advisory work has already had the pricing conversation. The profession has argued for years that hourly billing suits advisory badly, and most of the practices we speak with agree. They have moved to fixed fees, retainers, or tiered packages. The rate is gone.

The hours are not.

The short answer

A firm’s advisory capacity is set by what it costs to answer a question, not by how the firm charges for the answer. If each question requires reconstructing where the client stands, capacity scales with the number of questions, and the most available lever is limiting contact. That limit is usually expressed as a meeting schedule. Changing the pricing model does not move it. What moves it is making the client’s position standing rather than assembled on demand, so that much of the evidence an answer needs is already current when the question arrives.

Everyone agrees advisory should be continuous

This is not a contested point. Decisions do not wait for the calendar. A client signing a lease, taking on a hire, or accepting a large order with ninety day terms is making a finance decision in the moment, and the value of an accountant in that moment is high.

The published advice is consistent about this and also, in our reading, consistently incomplete. It establishes that time is the wrong basis for a fee, that pricing on outcomes rather than hours better reflects the expertise involved, and that a firm billing strictly by the hour may bill less for the same work when it becomes more efficient. All of that holds. Note that tracking time and billing by time are separate choices, and the measurement we propose below depends on tracking it. None of it tells a partner what to do on the Tuesday when nineteen client emails arrive.

That is the gap this guide is about. Not what to charge, but what has to be true for the work to be deliverable at all.

Why the meeting became the unit

Ask a firm how an advisory engagement is structured and the answer usually contains a number: monthly, quarterly, weekly for a few. Ask why that number and the answer is rarely about decisions. It is about what the engagement can carry.

The practical logic is sound. More contact means more preparation, more follow up, more of the partner’s attention, and that consumption is real whether or not anyone bills by the hour. A retainer set at four meetings a year has a knowable cost. A retainer set at “call me whenever” does not. So the interval gets fixed, and the fixed interval is what the client is told they are buying.

The result is that attention is still sold in countable blocks. The rate disappeared and the unit stayed. Moving to fixed fees does not remove the capacity constraint. Meeting frequency can become the way a firm contains preparation work, even when the client’s decisions arrive between meetings.

We want to be careful here, because cadence is not purely an artefact of pricing. Fit is real. A stable business with few decisions may find weekly contact intrusive, and at least one practitioner argument holds that meeting too often creates dependency rather than capability. Some clients genuinely need less. But when firms describe the constraint honestly, capacity is usually in the sentence somewhere, and that part is not about the client at all.

The repeated cost is assembling the picture

A substantial part of the cost comes before the partner can apply judgment: assembling the financial picture.

A client asks whether they can afford something. The partner knows the client, often well. What the partner does not have in front of them is the current position: where cash actually sits, what is already committed, what was decided last month and whether it happened, what the forecast expected and how far actuals have drifted from it. Some of that is in the accounting system, some in email, some in the partner’s memory, and some nowhere at all.

So the partner assembles it. That assembly is the real cost of a quick question, and it is the same work that makes call preparation take as long as it does. For questions about a new commitment, much of the preparation is required before the partner knows whether the answer will be straightforward. Routine lookups are different: they can often be answered directly from current records.

This is the actual capacity ceiling. Not the partner’s willingness, not the fee structure. The assembly cost, multiplied by the number of times someone needs an answer.

For the breakdown of what that assembly consists of month to month, and a way to measure it without double-counting, see where client advisory delivery hours go.

Why the usual remedies do not move it

Three responses are common and each one runs into the same wall.

Limit the contact. Batch questions to the scheduled call. This works, in the sense that it controls the cost, and it is what the meeting schedule is already doing. The price is that decisions get made without the firm, or wait, and waiting has its own cost.

Charge for the contact. Bill the questions, or add a support tier. Reasonable, and we have argued elsewhere that pricing the channel beats pricing the question. But pricing does not reduce the assembly cost. It funds it. The partner’s Tuesday is unchanged, and at some point the firm is out of Tuesdays regardless of what the engagement letter says.

Add people. Hire, or push work down to staff. This genuinely adds capacity and it is how firms have always grown. It is also the most expensive option, it requires the work to be legible enough to delegate, and controller-level attention tends not to form a visible queue that a firm can staff against.

None of these are mistakes. They are the available moves when assembly cost is taken as fixed. The question worth asking is whether it is.

What has to be true instead

If answering costs assembly, then the way to serve clients continuously is to make the position standing rather than assembled. Three things have to hold, and a firm can work toward all three without buying anything.

The client’s position is maintained between meetings, not built for them. Cash, commitments, and the current forecast stay current as a matter of routine rather than as preparation for a call. This is the largest change and the hardest, because most firms’ rhythm of work is organised around the call. The test is simple: if a client asks on a Tuesday, how much of the answer already exists?

Decisions and expectations carry forward. When a question arrives, it should arrive with its history. What was decided, what was expected to follow, whether it happened. Without that, every conversation re-establishes context that was established last month. A working commitment record is the mechanism, and it is buildable in a spreadsheet if that is what a firm has.

The client can see enough to answer their own lookups. A meaningful share of what reaches a partner is not a judgement question. It is where does cash sit, is that payment scheduled, what did we spend on this. Those consume assembly cost and return nothing that requires professional judgement. Giving the client a standing view of position and forecast removes much of that traffic, and it also closes part of what the client cannot see between check-ins.

The order matters. The third is the fastest capacity win and the one most firms skip, because it looks like a client service feature rather than a capacity intervention.

Does the upkeep cost less than the assembly?

This is the fair objection, and a partner should ask it before changing anything. Keeping cash, forecasts and commitments current is work too. Have you reduced the firm’s workload, or moved it from preparation into daily maintenance?

The honest answer is that it depends, and it is measurable.

Maintaining the picture creates capacity only when the upkeep takes less time than the repeated assembly it removes. Measure both: time spent keeping each client current, and time spent preparing answers. Include the exceptions that still require investigation.

Two things usually put the balance on the upkeep side, though neither is guaranteed. Maintenance is incremental where assembly is not, so keeping a forecast current from last week costs less than rebuilding it from the accounting system. And one maintained position serves every question that follows it, where assembly is paid again per question. The more often a client asks, the more the arithmetic favours maintenance, which is also why the clients worth starting with are the ones who contact you most.

Where it fails is the quiet client. A business that asks twice a year may not generate enough questions to repay daily upkeep. For that engagement, daily upkeep may not pay back through question handling alone. Set the maintenance cadence according to the client’s risks, commitments and agreed service. That is a real limit on this argument and worth knowing before a firm rolls the change across a book.

This is the measurement worth running for a quarter on two or three clients before committing. It also means tracking time, which is not the same as billing by it.

What this changes about cadence

It makes the interval a service decision again.

If the position is standing, a weekly client and a monthly client differ because their decisions arrive at different rates, which is the difference that should have been driving it. The meeting stops being the container for the relationship and becomes what it is better suited to be, which is the conversation where judgement gets applied to a picture both sides already have.

This is a real change in what a firm sells, and it is worth being honest about the size of it. It is more work up front than adjusting a fee schedule.

What a tool would have to do

Most software sold to firms produces an output: a report, a model, a screen to look at. That is useful and it is not the same as reducing assembly cost, because someone still has to decide what to look at and put the pieces together.

To move the capacity ceiling, a tool would have to maintain the position rather than render it. Keep cash and commitments current without being asked. Hold what was decided and what was expected, and check whether it happened. Notice when actuals have drifted from the forecast and say so before someone asks. Give the client a view they can use themselves. The test is not whether it produces a good answer when queried. It is how much of the evidence needed for an answer is already current and assembled when the question arrives.

Five questions to ask about your own practice

1. When a client asks a question on a Tuesday, how much of the evidence the answer needs is already current, and how much has to be assembled?
2. What share of the questions you receive are lookups rather than judgement calls?
3. Is the meeting interval on your largest engagements set by how often decisions arrive, or by what the engagement can carry?
4. If a client made a decision three weeks ago, would you know today whether it happened?
5. What would have to be true for you to tell a client to call you whenever, without the cost of that being unbounded?

Our product

AgentLink (agentlink.finance) is built to hold the position rather than produce a report: cash and commitments kept current, decisions and expectations carried forward, drift from the forecast flagged before it is asked about, and a client-facing view for routine lookups. It does not remove the judgement, the investigation a novel question needs, or the exceptions. It reduces how much has to be rebuilt first. It is sold through CPA and fractional CFO firms. The firm keeps the client, the pricing and the judgement. Nothing in this guide requires it.

Frequently asked questions

Is this an argument against value pricing?

No. Pricing on outcomes rather than hours is the right direction and most of the firms we speak with have already moved. The point is that it solves a different problem. Value pricing changes what the client pays for. It does not change what it costs the firm to answer, and that cost is what sets capacity.

We already offer unlimited support in our top tier. Is that not the same thing?

It is the right commercial shape and it does not by itself reduce the work. Unlimited support with assembly-heavy answering means the firm has priced in a cost it has not reduced. It works while client count is low and gets harder as it grows.

Does this mean meeting less often?

Not necessarily, and often the opposite. It means the interval stops carrying the whole relationship. Some clients will still want a scheduled conversation at the same frequency, with the difference being what has already happened before it starts.

How is this different from just keeping the books current?

Current books are necessary and not sufficient. Books record what happened. Answering whether a client can afford something also requires what is already committed, what the forecast expects, and what was decided and did not happen yet. That is a different set of facts than a clean reconciliation.

What is the smallest version of this a firm can start with?

Give clients a standing view of cash position and the near-term forecast, so fewer routine lookups reach the partner. It is the least disruptive of the three and usually the largest immediate reduction in traffic.

Is this not just moving preparation work into daily maintenance?

It can be, and that is the test to run before committing. Maintenance is only a gain when the upkeep costs less than the repeated assembly it removes, which depends on how often a given client generates questions. Measure both for a quarter on a few clients rather than assuming either way.

Does this apply to compliance work?

Less so. Compliance work has a calendar for good reason, and deadlines are external. This is about advisory work, where the schedule is chosen rather than imposed.

About the author

Nilanjan Raychaudhuri is the founder of AgentLink (Tublian LLC, Columbus, Ohio), which builds controller-layer software sold through CPA and fractional CFO firms. He has spent the past year interviewing practitioners at fractional CPA and CFO practices about how advisory work is actually delivered. Team page

The practitioner observations in this guide come from design-partner conversations and are used with permission where attributed. Corrections to nilanjan@agentlink.finance.