How to measure the impact of your advice

The three earlier guides in this series covered the mechanics of a client check-in: what to review beforehand, how calls go wrong, and how long the preparation takes. None of them asked why any of it matters. This one does.

A firm prepares for a call, gives advice, and the client leaves with a handful of decisions. Three months later, can the firm say which of those decisions were carried out, what happened as a result, and whether the advice was any good? For most of the practices we spoke with, the honest answer was no. Not because the advice was weak, but because nothing in the firm’s toolkit records what became of it.

The short answer

Advisory impact cannot be measured from the books, because the books record transactions and not decisions. To measure it, a firm needs a record that holds each recommendation, what the client agreed to do, what observable result should follow, and whether it did. Once that record exists, three things become possible that are not possible today: the firm can see which advice was acted on, it can see whether the intended result followed, and both the firm and the client can stop relying on memory to settle what was said.

One limit needs stating up front. A record can measure whether advice was accepted, whether the agreed action occurred, and whether the intended result followed. It cannot, by itself, prove that the advice caused the result. Market conditions, other management decisions and timing can all produce an outcome the advice was meant to produce. That final assessment remains professional judgement. What the record does is give that judgement something to work from other than memory.

That last one is the part most firms underestimate. Measurement and accountability are the same mechanism. A record that lets you score your advice is also the record that keeps both sides honest about it.

Advisory is defined by its output, not its outcome

The AICPA defines advisory services as work where the practitioner develops findings, conclusions, and recommendations for client consideration and decision making. The definition identifies findings, conclusions and recommendations as the practitioner’s output. It does not provide a method for recording what the client accepted, whether the agreed action occurred, or what followed. How far an engagement extends into implementation depends on its scope, and many advisory engagements do extend that far.

But the definition shapes the artefacts. Because the output is defined as the recommendation, the natural artefacts of advisory work are built around the recommendation: the memo, the deck, the meeting summary, the follow-up email. All of them capture what was said. None of them are built to capture what happened next, even in engagements where what happened next is squarely the firm’s concern.

Practitioners who have thought hard about advisory tend to notice this. One Accounting Today column on what advisory actually means argues that firms need to develop metrics to measure the before-and-after impact of their advice. We agree. The question this guide takes up is why that is harder than it sounds, and what it takes to do.

The reports answer state, not consequence

When we raised this with practitioners, the most common objection was some version of: I already have the reports. They tell me everything about the company.

They do, within a limit. Financial reports describe the recorded state of the business, subject to their close and reconciliation status. A current set of financials, a bank feed, an aging report and a cash position answer the question “what is the recorded state of this business right now,” and where the books are reconciled and categorised they answer it well. Nothing in this guide argues otherwise.

What the reports cannot show is how that state relates to advice given months earlier. In March the adviser recommended holding the owner distribution until the receivable from the largest customer cleared. Did the client hold it? The bank statement shows a transfer on the fourth of April. Was that the distribution, made anyway? A smaller one, as a compromise? Something unrelated? The transaction is right there. Its meaning is not.

Multiply that across the two to five decisions that come out of a typical call, across a book of fifteen or thirty clients, over a year. The reports describe the business faithfully every month. They record no decision, no recommendation, and no commitment, so they cannot say which of the firm’s advice was taken, let alone whether it helped.

This is why preparation feels like work that is not work. Much of what fills the hours before a call is not reviewing the business; the reports do that quickly. It is reconstructing the thread: what did we agree, what were they going to do, did they do it. Experienced advisers get faster at this. They do not get to skip it, because the thing they are reconstructing was never written down in a form that could be checked.

Firms do not know which advice was acted on

There is no reliable published estimate of how often accounting advisory recommendations are implemented. We looked. The nearest figures come from management consulting, rest on small informal surveys and individual consultants’ recollections, and say nothing directly about CPA or fractional CFO practices, so we are not going to repeat them here.

That absence is itself the finding. Firms remember examples of advice that worked, and usually the ones that did not, because those come up. What no firm we spoke with could produce was its own rate: of the recommendations made to this client over the past year, how many were acted on, and of those, how many produced the intended result. Not because the firms were careless, but because the number cannot be calculated from memory and email, and nothing else was holding the inputs.

An accounting advisory relationship should be unusually well placed to know this, because the adviser is still there next month. The relationship is continuous; the record is not. A firm that meets a client twelve times a year has twelve chances to notice that a decision quietly did not happen. Without a record, most of those chances are spent re-establishing what was decided rather than checking on it.

Accountability runs both ways

Advisers are sometimes uneasy about recording client commitments, because it can feel like building a case. It is worth being direct about why the record protects the adviser at least as much as it holds the client to account.

Without a durable record, both sides reconstruct the relationship from memory, and memory is self-serving on both sides. When something goes wrong, the client’s version is that the firm should have caught it. The firm’s version is that it was raised in March and the client chose not to act. Both people believe their version. Neither can prove it. The conversation that follows is about who is right about the past, which is the least productive conversation a firm can have with a client it wants to keep.

A record ends that conversation before it starts. Not by assigning blame, but by removing the need for either party to be right from memory. The recommendation was made on this date. The client agreed to this. The observable result was this. It happened, or it did not, on this date. Nobody has to remember. Nobody has to argue.

This is what accountability actually looks like in a professional service. It is not the firm holding the client accountable. It is both parties being accountable to the same record, one that the firm keeps because the firm is the party being paid for judgement, and judgement that cannot be checked afterwards is indistinguishable from opinion.

Why the economics of the service depend on it

Advisory is priced above compliance because it is supposed to change outcomes. Wolters Kluwer puts advisory fees at around five times traditional tax preparation fees on average, and the case for value-based pricing rests on being able to point at outcomes rather than hours.

A firm that cannot show which advice was taken and what followed is pricing on outcomes it cannot demonstrate. That holds up for as long as the client trusts the relationship, which is often a long time. It does not hold up at renewal when a cheaper alternative appears, or when the client’s business hits a rough patch and starts asking what the advisory fee has bought.

The firms that will hold their pricing are the ones that can answer: here are the fourteen decisions we worked through this year, here are the eleven you acted on, here is what happened on each, and here are the three we would raise again. That is a measured service. It is also a service that gets better, because the record gives the adviser a history of what they recommended and what followed, which is the raw material for judging where their own judgement has held up.

The same record is what stops the same three questions from being re-derived every month. Where should we focus. What has changed. What should we do now. Without a record of what was decided last time, each call answers those from scratch. With one, each call continues the previous answer, and the firm’s judgement compounds instead of resetting.

What a measurable advisory record requires

The mechanics of keeping the record are covered in the guide on tracking client commitments, and we will not repeat them here. What matters for measurement is that the record holds five things for each recommendation.

The recommendation as made, with its date. Not the polished version from the follow-up email, but what the adviser actually said the client should do, and when.

What the client agreed to. This is often narrower than the recommendation. The adviser suggests holding the full distribution; the client agrees to hold half. Record the agreement, not the advice, because the agreement is what gets scored.

The observable result that should follow. A decision with no observable result is a discussion, and discussions are not measured. “Reconsider pricing” produces nothing to check. “Revised rates appearing on invoices to the top ten accounts from the first of next month” does.

Whether it happened, verified against the bank or the books where financial evidence exists, and otherwise confirmed explicitly with the client and marked as unverified. This is the field that turns a note into a record, and it is the one that cannot be filled in from the record itself.

What followed. Did holding the distribution avoid the shortfall it was meant to avoid? This field records the observed result. It does not, on its own, establish that the advice produced it.

Once these exist across a client’s history, follow-through and outcome rates become arithmetic. The share of recommendations acted on. The share of those where the intended result followed. The kinds of advice a given client tends to take and the kinds they tend to let go. Assessing how much the advice contributed to each result remains judgement, but it is judgement applied to a record rather than to a recollection. None of this requires a system. It requires the record to exist, and someone to fill in the fourth and fifth fields by looking at what actually happened.

Where this leaves preparation

Read the three earlier guides again with this in mind. The checklist of what to review is mostly state recovery, which the reports handle, plus thread recovery, which nothing handles. The failure modes are largely what happens when thread recovery is incomplete. The time preparation takes is, in large part, the time thread recovery takes when it has to be done from memory and email.

A firm that is excellent at preparation under these conditions is not performing well so much as compensating well. The compensation is real skill and it is worth having. But it is compensation for the absence of a record. The record does not stand in for that skill. It preserves the judgement the skill produces, makes it available across the firm rather than inside one person’s memory, and removes the reconstruction work that surrounds it.

The reason to prepare well is not the call. It is that the call is where advice gets given, and advice that is never checked afterwards cannot be measured, cannot improve, and cannot be defended. Preparation is the front half of a loop whose back half most firms have never closed.

Our product

AgentLink (agentlink.finance) keeps this record on the firm’s behalf. Decisions from each call are captured as commitments, with what was agreed, the observable result, and the promised date. Bank and book activity is checked against them daily, so the question of whether each commitment happened is answered before the next call rather than on it, and the open items are brought back to the adviser for confirmation. The firm keeps the client, the pricing and the judgement. AgentLink supplies the attention between calls.

Frequently asked questions

Isn’t recording client commitments a liability risk for the firm?

Clear, contemporaneous documentation can reduce ambiguity about what was recommended and what the client decided, but it does not eliminate liability risk. Risk-control specialists at CNA, the underwriter of the AICPA professional liability programme, have described deficient documentation as a pattern present in large claims, noted that this holds across tax and consulting work and not only audit, and made the point that when there is a dispute your documentation tells your story, and gaps in it get filled in by someone else. The record should match the engagement’s defined scope, distinguish the firm’s advice from management’s decisions, and follow the firm’s retention, consent and privacy policies. Review the approach with your professional-liability adviser or counsel before adopting it.

Our clients don’t want to feel monitored.

Practitioners raised this, and the answer that worked in practice was framing. The record is not surveillance of the client. It is the firm’s own account of its advice, kept so the firm can show what its work has done. Most owners, told that their adviser keeps track of whether the advice worked, hear diligence rather than suspicion.

How is this different from the commitment tracking guide?

That guide covers how to keep the record week to week. This one covers why. The commitment record answers "did it happen." Measuring advisory impact adds "did the intended result follow," and uses the accumulated history to inform the firm’s judgement about its own advice, not only the client’s follow-through.

What if the client took the advice and it still went badly?

Then the record shows that too, and it is the most valuable thing it can show. Advice that was followed and failed is where an adviser learns. Without the record, that failure gets attributed to the client, to the market, or to nothing, and the same recommendation gets made again.

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.