What your monthly reports do not answer

Most small businesses should receive three core statements every month: profit and loss, balance sheet, and cash flow. Add receivables and payables aging, payroll liabilities, and a note on anything unusual, and you have the standard package. Monthly is right for most businesses. Quarterly is defensible for the quietest ones. That is the honest short answer to the question most owners are asking when they search for it, and a great many firms have written it up well.

It is also the answer to a narrower question than the one owners actually have.

The monthly package describes performance over a completed period and financial position at its reporting date. It does not automatically reflect changes since then. That makes it a good record of what happened. It is not an answer to where the business stands today, what is already committed against the cash currently in the account, or whether a decision being considered on the fourteenth is affordable. Those questions do not arrive on a monthly cycle. They arrive whenever they arrive, which is usually at the least convenient moment, and the honest current answer is that the owner asks and waits.

So the split worth drawing is not between which reports to send and how often. It is between what genuinely needs a person to assemble, interpret and stand behind, and what is a lookup: something that needs to be correct and current and framed properly, but needs nobody’s judgement at the moment it is asked. Most firms have never separated the two explicitly, so everything lands in the same queue and gets the same treatment.

Drawing the line is worth doing, and it is a capacity decision before it is a service decision.

What the client is actually buying

Ask a business owner why they hired a firm and very few of them describe a reporting package.

They say they wanted someone who knows the business. Someone they can ask. Someone who will say something if a number starts moving the wrong way, rather than waiting to be asked. Someone who cares whether the business works.

The monthly reports are a proxy for all of that. They are the visible, billable, schedulable evidence that somebody is paying attention. When a client says the reports feel late or thin, they are usually not making a claim about the reports. They are saying the relationship feels further away than they expected it to.

This matters when a client complains about the reports. Some of those complaints are literal and worth taking at face value: the package really is late, or missing a schedule, or hard to read. Others are about the format only in the way that a locked door is about the handle. Adding a chart to the package does not address a client who wanted to be able to ask.

Why that is hard to deliver, and why it is not unwillingness

Firms are not withholding. The structure of the work makes continuous contact expensive in a way that is easy to miss from outside.

Questions about commitments cost assembly rather than lookup. Answering “can we afford this hire” means pulling current position, checking what is already committed, remembering what was agreed two calls ago, and reconstructing enough context to say something responsible. That reconstruction is the cost, and it recurs per question because the assembled version is not kept anywhere. We worked through the mechanics of this in what a quick question actually costs your firm.

Because each contact carries that cost, firms batch. Batching is the rational response: assemble once a month, deliver everything at once, amortise the reconstruction across everything the client wanted to know. The monthly cadence is not a service philosophy. It is what the cost per contact produces when you optimise it.

The client experiences the batching as distance. They also experience something more specific, which is that between packages they can see their bank balance and their own commitments but cannot see the two combined. What your clients cannot see between check-ins covers that gap in detail.

And the cost per contact is what caps how many clients a practitioner can carry at all. That is the argument in how to scale advisory work without scaling hours and, for a single practitioner’s client load, in how many clients a fractional controller can carry.

So the split between what arrives monthly and what is checkable any time is not primarily about being more generous with clients. It is about which contacts require a person and which ones only require that something correct already exists. Every question that moves from the first category to the second is capacity returned to the firm.

The usual answer solves a different problem

Search for how firms give clients access and you will find portals. Dozens of them, and the comparisons between them are genuinely useful if you read what they compare on: encryption, document collaboration, white labelling, role based permissions, two factor authentication, audit trails, electronic signatures, request lists, payments.

Every one of those criteria is about moving documents safely. That is a real problem. Tax documents have to get from the client to the firm, engagement letters have to be signed, sensitive files should not be sitting in email. A portal is the right shape for that, and firms that have one should keep it.

But a document portal is a place files go. What we are describing is a place a shared position lives. Those are different objects. A folder of last month’s statements answers “where is the thing you sent me.” It does not answer “where do we stand right now,” because a delivered file reflects its reporting date rather than the present and because the answer to that question was never in the statements in the first place.

The distinction is worth being precise about, because firms evaluating this often think they have already solved it. If the thing you have is a well organised set of documents, you have solved filing. The client who wants to know whether they can afford the hire still has to ask.

Three categories, not two

Here is the part worth being concrete about. What follows is a starting split, not a standard, and firms will move items between the categories based on the client and the engagement.

The common mistake is to sort everything into two piles: what the firm sends on a schedule, and what the client can see for themselves. That collapses two separate questions into one. Whether something requires the firm’s judgment is a different question from whether it should arrive on a cadence. Some things need judgment and cannot wait for the next package. Some things need no judgment at all but are still worth pushing rather than leaving to be found.

Three categories hold the distinction properly.

The organising principle across all of them: what a client can see has to carry enough context to be interpreted correctly on its own, at eleven at night, with nobody available to ask. If a number cannot survive being read alone by someone who is tired and worried, it does not belong on the self-serve side, no matter how current it is.

One: delivered on a schedule

Assembled by the firm and pushed to the client at a fixed cadence, whether or not the client asks. These are periodic by nature because they describe a period.

• The period’s results, with commentary on what actually happened rather than only what the numbers were.
• Performance against the approved plan, with the explanation for material variances. The variance is data. The reason is the work.
• Routine deadline status: what has been filed, what is on track, what is coming up in the ordinary course.
• A summary of what changed since the last package and what the firm did about it.

Two: available for the client to check

Current, correctly framed, and requiring nobody at the moment it is read. These are not periodic. Making them periodic is what creates the gap the client experiences as distance.

• Current cash position, and what is already committed against it. Not the bank balance, which the client already has and which is the number that misleads them.
• The forecast. Specifically, the same forecast the firm is working from, not a client facing copy of it. Two versions of the forecast is the failure mode this is meant to prevent.
• The approved annual plan, sitting alongside that forecast, so any number is read against what the business intended rather than in isolation.
• The record of decisions: what was decided, when, what the plan said at the time, what the forecast said, and why the business proceeded. Without this a moving forecast is unexplainable. The client sees a number change and cannot tell whether reality moved or whether it reflects something they agreed to three weeks ago.
• Open items and their status. What the firm is waiting on, what the client is waiting on.
• The state of the underlying data. When bank feeds last refreshed, when accounts were last reconciled, and the total amount of activity not yet classified, with a plain statement of how those gaps affect the figures shown.

Two of those deserve emphasis because they are the ones firms hesitate over.

The same forecast, not a second version of it. The distinction that matters is between presentation and substance. Firm and client should be working from the same reviewed numbers and the same assumptions. Their views of those numbers can and usually should differ in detail: the firm’s working view carries the line level build, the client’s view carries categories and the assumptions stated plainly. That is a presentation choice and it is fine. What is not fine is a separate set of numbers or a separate set of assumptions, because then each side holds a different picture without knowing it and the first twenty minutes of every call goes to reconciling them.

The plan alongside it. This is the guardrail that makes exposure safe. A number on its own invites a conclusion. A number against a plan invites a question, which is the response you actually want, and the question arrives at the firm already framed.

What does not belong in this category, and why:

• Unclassified transactions, item by item. Not because they are secret but because a client cannot tell which ones are noise. Their financial effect should still be visible: the total amount awaiting classification, and which figures it could move, stated alongside the figures themselves. Categories in the client’s own language rather than raw line items is a safety property, not a design preference. Withholding the size of what is unresolved is not.
• Anomalies that have not been triaged yet. A flag with no assessment attached is an alarm, and an alarm at eleven at night produces a phone call, not a decision.
• Work in progress. Draft adjustments, reclassifications mid review, anything that will be different by morning.

Three: requires the firm’s judgment

This is the category most often folded into the first one, and folding it there is the error. These questions need a person, but they do not fit the reporting cycle. Questions requiring action before the next reporting date need an agreed response path between scheduled reviews.

• Whether a proposed commitment is affordable, and what it displaces if it is taken.
• Material risks, and approaching deadlines that require action rather than awareness. These do not wait for the next package.
• What a change in the outlook actually means for the business, as distinct from the fact that the outlook changed.
• What the firm recommends doing about anything in categories one or two.
• Any assessment the firm has not yet formed a view on. If the firm has not decided whether something is a problem, exposing the input without the view asks the client to form the view instead, which is precisely the expertise they are paying not to need.

The point of getting category two right is that it shrinks the queue in front of category three. Every lookup that no longer arrives as a request is capacity returned, and the questions that remain are the ones worth a person’s time.

None of this is gatekeeping, and it reads as gatekeeping only if category one is thin. A client who receives real interpretation every month does not experience “let me look at that properly and come back to you” as a brush off.

If you are the business owner

You can ask for this. Most owners do not, because the menu they have been shown contains report formats and delivery dates, so those are the things they negotiate.

The more useful conversation is about which of your questions actually need your accountant’s judgement and which ones are you asking only because you have no other way to find out. Sort your own questions that way before the next conversation. The lookups are the ones worth raising, because they are the ones where something could exist that answers them without costing anybody a call.

Talk to your CPA about the possibilities.

If you are the firm

The decision is not whether to give clients more access. It is which contacts require a person.

Take the questions you answered for one client over the last quarter. Sort them into ones where the value was your judgement and ones where the value was that you knew where to look. The second pile is the specification. Everything in it is a contact you are currently paying assembly cost for and getting no advisory credit for.

Then decide what the client sees and, more importantly, what it sits next to. Nearly every risk in exposing financial information to a client is a context problem rather than an access problem. A cash number alone is dangerous. A cash number against a plan, with the decisions that moved it visible, is the beginning of a better conversation than the one you would otherwise have had a week later.

Common questions

Is this not just a client portal?

A document portal solves secure file exchange, which is a real and different problem. This is about whether a current shared position exists that both sides read from. A firm can have an excellent portal and still have every client question arrive as a request for assembly.

Will clients misread the numbers and panic?

Some will, if they are given numbers without context. That is the reason for pairing the forecast with the approved plan, bucketing by category instead of line item, and keeping untriaged anomalies off the client side. The failure mode is not access, it is a number that cannot be interpreted alone.

Does this mean clients stop calling?

No, and it should not. It changes what the calls are about. Fewer lookups require a call. The calls that remain are the ones where the client wants your view, which are the ones worth having and the ones you can charge for.

Should every client get the same split?

No. A client with a single revenue line and predictable costs needs less available continuously than one carrying inventory and seasonal swings. The lists above are a starting point to adjust from, not a standard to apply uniformly.

What if the working forecast is not in a state to show a client?

Separate the two things that question usually bundles. If the issue is presentation, that is solvable: the client’s view can show categories and stated assumptions while the firm’s working view keeps the line level detail. If the issue is that the numbers or the assumptions have not been reviewed, that is a different problem, and it affects the advice the firm gives as much as anything the client would see.

Does the client see the decision record?

That is a firm choice, and it is covered in more detail in tracking client commitments. Our view is that a forecast without the decisions behind it cannot be explained, so if the forecast is visible the decisions should be too.

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.