What a quick question actually costs your firm

One practitioner we work with described his working day this way: it is decided by whatever is in the inbox when he opens it. The monthly check-ins are the only fixed ritual. Everything between them is set by whatever clients sent overnight, and a good share of that is the message that begins with “quick question.” Can I bring this hire forward. Can I pay the vendor early for the discount. Is the new lease fine. We have heard versions of this from nearly every firm we have spoken to.

The profession has a name for this and a standard answer. The name is scope creep. The answer is an engagement letter that draws the line, a work order for anything past it, and time tracking so the leak is visible. All of that is sound, and this guide does not dispute the work is real and mostly unpaid. It disagrees about two things: what the question actually costs, and whether the standard fix is worth what it gives up.

The short answer

A quick question is expensive because it asks about a commitment, and the books show a position. Answering well means assembling four things: where the business stands now, what the forecast expects, what the plan intended, and what has already been decided since the last call. Each is quick to fetch. Nothing holds them together, so the assembly happens fresh every time, and the fourth item is often scattered across notes, messages and memory. The cost is not looking anything up. It is putting the picture back together before you can think, and the context switch that comes with it.

Where scope control means a separate fee discussion for every small question, it can discourage the early conversations that make advisory useful. A client who learns that asking has a price may stop asking and decide on feel instead. The loss from a decision made on gut that could have been made on numbers is larger than the unbilled twenty minutes, and it tends to arrive back at the firm later as a problem that was preventable. And even a well-priced support channel still carries the assembly cost, which is why this guide spends most of its length there.

Why it costs more than it looks like it should

Start with the case where everything is right. The books are current. Reconciliation is done. The forecast was rebuilt at the last close and is a good one. The client asks whether they can add a second technician in October.

If the partner has that client in their head, this is a five-minute answer. Most partners at a twenty- or thirty-client practice do not have any single client in their head on a given Wednesday. So they look. And what they can look at, quickly, is a report. A report shows position: cash, receivables, payables, run rate. The question is not about position. It is about whether a specific new outflow fits against the position, the expectation, and the plan, given what else is already committed.

That takes four sources.

Current position. The fastest of the four. If the books are current, this is a lookup.

The forecast. Also fast if one exists, and slower if the last one is from the previous close and the month has moved.

The plan. The budget, if there is one. What the year intended, and where the lines under pressure are. Many owner-led businesses do not have this in formal form, which makes it a memory exercise rather than a lookup.

Decisions already in flight. This is the one that does the damage. The client agreed to a marketing spend on the last call. They mentioned a price increase they were thinking about. They said the equipment lease was likely. None of that is in the books because none of it has happened. It is in the partner’s notes, in the partner’s memory of the call, or nowhere. A hire that looks affordable against the bank balance and the forecast can be unaffordable once the commitments from three weeks ago are laid on top, and the report cannot say so. The record needs to distinguish proposed decisions from approved commitments, and show which commitments are already included in the forecast, because a commitment the forecast already carries must not be counted a second time.

This is the gap the budget series named from the owner’s side: the budget, the forecast, and the outstanding commitments are rarely in one place when the decision arrives. This is the same gap from the firm’s side, with a price tag on it.

Illustrative calculation. Twenty questions a week across the practice, at fifteen minutes of assembly each, is five hours a week before judgement and before writing the replies. Those figures are illustrative, not measured. Substitute the numbers from your own question log, which the pricing section below explains how to keep.

The client you have not touched in five weeks

Assembly cost is not evenly distributed. The client whose call was last week is cheap to answer because the picture is still warm. The client whose call was five weeks ago is the expensive one, because the fourth source has degraded, the forecast has aged, and the position has moved in ways the partner has not seen.

For the client whose last call was five weeks ago, the partner has more to reconstruct before answering confidently. The client sees the same short reply. The firm absorbs the difference in preparation time, and because that time is spread across an inbox rather than a project, it never shows up anywhere as a cost.

What the standard advice gets right, and where it goes wrong

The advice is right that the work is real. A firm that answers dozens of these a week without any mechanism to see it is subsidising advisory work it never priced.

The advice is not uniform about the fix, and it is worth being fair about that. The value-pricing school has argued for a long time that access to questions belongs inside a fixed-price agreement rather than on a clock; a 2008 Journal of Accountancy feature profiles a firm whose agreements include unlimited access precisely because clients resented being billed for a ten-minute call. That is the right instinct, and this guide builds on it rather than against it.

Where the common remedies go wrong is when scope control requires a separate fee discussion for every small question. Batch it for the monthly call. Quote before answering. Point to the engagement letter. Each of these works as designed, and the design is friction at the point where the client is trying to make a decision. Batching adds delay to a question that has a deadline. Per-question pricing puts a meter on the interaction where the client is most exposed to a bad call. Both succeed by reducing the number of questions, and that is the cost.

What happens when the client stops asking

The client still has to decide whether to make the hire. If asking has become slow or awkward or expensive, they decide anyway, on feel. Owners are good at feel. It is how they got to this size. It is also how they overcommit in September on the strength of a strong August, and how they decline a hire that the numbers would have supported.

Two things follow for the firm. First, the loss lands back on the firm later, because the consequences of a decision made without the numbers show up in the numbers eventually, usually as a fire drill. Second, and harder to see, a client who asks freely is a client who is close. A client who has gone quiet may simply have had a quiet quarter, or may be making decisions elsewhere. Silence does not tell you which, and that is the point: without a record of the questions, the firm has no way to notice the change. The question volume was never only the leak. It was also the pulse.

Price the channel, not the question

Per-question pricing prices the wrong unit. It inserts friction where friction should be lowest, and it makes the client feel metered for the thing they most value.

The alternative is to price the channel. The retainer is built as two explicit parts: the scheduled work, and a named continuous support channel with an allowance attached. Not a single number that silently absorbs the questions, but a line the client can see. The tiers most firms already run do this implicitly, in that the weekly client pays more than the monthly one. Making it explicit changes what happens when a client’s question volume runs high. Instead of feeling like leakage, it becomes the reason for a tier conversation, and that conversation is easier because the client can see what they have been using.

This only works if the firm measures it. Log the questions: who asked, when, roughly what it was about, and how long a good answer took. Nothing more elaborate than that. Over a quarter the log shows which monthly clients are effectively already weekly, and moving them up becomes something you show them rather than something you sell them. The same log shows the client who has gone quiet, which is the retention signal no other report produces.

We offer this as something practitioners should consider, not as pricing guidance. Retainer structure depends on the practice. The structural point is narrower: leaving the support channel unnamed and unmeasured is what makes it feel like a leak, and what makes the only available fix a meter.

What reduces the cost of answering

Pricing the channel does not make the answer cheaper to produce. Two things do.

The index document. One page per client, refreshed at close, carrying the four things the answer needs: the current position in a few numbers, the shape of the next quarter’s forecast, the budget lines under pressure, and the decisions that are open or recently made. It does not answer any question. What it does is collapse the assembly. The partner reads one page instead of gathering four things, and the fourth thing, the in-flight decisions, has a home. Its limit is honest: it goes stale between closes, and the client you have not touched in five weeks is still the one whose page is least reliable. But a stale page beats no page, because it at least records what was in flight the last time anyone looked.

A self-serve channel for the routine. A meaningful share of quick questions are lookups dressed as decisions: what did we spend on that, is the tax payment scheduled, where does cash sit today. Those should never reach the partner. A client-facing view of position and the forecast, which the earlier guide on what the client cannot see covers, takes the lookups off the inbox and leaves the questions that need judgement.

Large firms solve the whole problem a different way: with people. A dedicated engagement team and a named contact who is expected to respond promptly, with coverage when that person is out. That is headcount, and it is affordable because the client pays enough for it. It does not scale down to a twenty-client fractional practice where the same inbox lands on one partner.

What a tool would have to do

If software were to take on the assembly rather than the answer, the requirements follow from the four sources. It would need to hold the current position from the bank, not from a report that lags it. It would need to carry the forecast and the plan as separate things, with the plan held as the firm supplied it. And it would need a place for the fourth source: the decisions and commitments from calls, entered once and kept current, so that “can I afford this” is tested against what is already committed and not only against the balance. It would need to keep that picture current between closes, because that is exactly when the question arrives. And it would need to say when it cannot answer, rather than answering anyway.

None of that supplies judgement. The judgement is still the partner’s. What it supplies is the assembled picture the judgement needs, at the moment the question is asked rather than at the next close.

Five questions to ask about your own practice

1. When a client asked an affordability question this month, how many places did you look before you could answer?
2. Was anything from the last call relevant to that answer, and where was it written down?
3. Which of your clients would you answer least confidently today, and when was their last call?
4. How many quick questions did your practice answer last month, and does any record say so?
5. Which client has not asked you anything in the last two months?

Frequently asked questions

Is this just an argument for not billing for advisory time?

No. The work is real and should be paid for. The argument is about the unit. Price and measure the channel rather than the individual question, so the client is never made to hesitate at the moment of a decision.

What if the client's books are not current?

Then the catch-up is work that already existed, pulled forward by the question. It is a real cost, but it is not the cost of the question, and firms that keep books current still pay the assembly cost this guide describes.

Would a shared Slack channel or client portal fix this?

It reduces friction to ask, which is worth doing. It does not by itself reduce the cost to answer, because the answer still requires the four sources to be assembled. A portal that only shows documents and position takes the lookups off the inbox and leaves the decisions where they were.

How is this different from the scope-creep advice already out there?

Some existing advice already includes ongoing access within a fixed fee. This guide builds on that approach by addressing the cost that remains: assembling the financial picture each time a question arrives. Pricing the channel pays for the work; keeping that picture current reduces the work.

What is the fourth source, exactly?

The decisions and commitments that have been made or discussed since the last close and have not yet shown up in the books. The earlier guide on tracking client commitments describes the record.

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.