Most monthly financial reporting stops at the first question. It tells management what happened. Revenue was $89.3K, up 7 percent on prior year. Gross margin held at 85.5 percent. Operating expenses were $18.5K.
All true. None of it decisions.
A management briefing is a different artefact. It is built on the observation that management does not need to be told the numbers — they need to be told what the numbers mean and what to do next. That requires carrying each observation through a chain of reasoning that a dashboard is not designed to hold.
Every operating area in a briefing answers the same four questions, in the same order:
| Station | Question | What it produces |
|---|---|---|
| 1 | What happened? | The figure, the variance, and whether it sits inside or outside a relevant benchmark range. |
| 2 | Why did it happen? | A decomposition into drivers that sums exactly to the total change. |
| 3 | What happens next? | How much of the forward period is already committed versus modelled. |
| 4 | What can we do about it? | Quantified levers with their assumptions disclosed. |
The consistency is not cosmetic. When every area follows the same structure, management learns to read the document once and then navigates it without help. More importantly, they begin to reason the same way about problems the briefing does not cover.
Question one is easy: it is a number and a comparison. Question two is where analysis begins, and it is where most reporting substitutes description for explanation.
"Revenue was down due to lower sales" is a restatement, not an explanation. A real answer to question two decomposes the change into named components that sum to the total:
When those five components are quantified, the conversation changes. "Revenue grew 7 percent" becomes "growth is expansion-led at $12.7K, against churn of $14.8K — the base is shrinking while existing customers spend more." Those are different businesses requiring different responses, and only the second framing tells you which one you are running.
The reconciliation test: if the components of your explanation do not sum exactly to the change you are explaining, the explanation is incomplete. Something is missing or double counted.
Most monthly packs treat the forecast as a separate document produced on a separate cycle. That is a mistake, because the most useful thing management can know about a variance is whether it persists.
Question three asks how much of the forward period is already locked. For a business with deferred revenue schedules, a meaningful share of next quarter is contractually committed and already sitting in the accounting records. Separating committed revenue from modelled revenue changes how management reads a bad month: a shortfall against a 93 percent contracted base is a different problem from a shortfall against a fully modelled forecast.
The same applies to cost. Signed hires, loan and lease payments, and renewal commitments are known. Presenting them as committed rather than forecast tells management what is genuinely discretionary.
Analysis without a recommended action is homework. Question four asks what management can actually pull this month, and it demands quantification.
"Consider improving collections" is not a lever. "Extending supplier terms by 15 days releases $16K of cash on a one-time basis" is a lever, because it has a magnitude, a mechanism, and a disclosed basis that someone can challenge.
The discipline here is that every lever must state its assumption. A recommendation whose basis cannot be shown is an opinion, and management should treat it as one.
A margin decline of 200 basis points means something different in a period of 3 percent inflation than in a period of zero inflation. A DSO of 66 days means something different depending on industry norms.
Including reference points — central bank policy rates, consumer price inflation, and industry benchmark ranges for gross margin, EBITDA, and days sales outstanding — lets management judge whether a result is company-specific or market-wide. That distinction usually determines whether the correct response is operational or strategic.
The requirement is that external data is labelled as external and its source and vintage disclosed. Benchmark ranges drawn from a curated pack should say so, rather than implying live market data.
It is not a longer dashboard. Adding narrative text to a metrics display does not produce a briefing if the narrative is descriptive.
It is not a board deck. A board deck is a presentation artefact with a persuasive structure. A briefing is a reasoning artefact with an analytical structure. The board deck can be assembled from the briefing, but they are different documents with different jobs.
It is not AI-generated commentary. Generated text can produce something that reads like a briefing without any of the underlying decomposition. The distinguishing feature of a briefing is that every claim can be traced back to the accounting records that produced it.
A management briefing is a structured monthly document that answers four questions for each key area of financial performance: what happened, why it happened, what happens next, and what management can do about it. Unlike a dashboard, which displays metrics, a briefing carries the reasoning from observation through to recommended action, with each figure traceable to the underlying accounting records.
A dashboard is a display layer. It shows current values, trends, and variances, and leaves interpretation to the reader. A briefing is a reasoning layer. It decomposes each variance into its drivers, connects those drivers to the forecast, and quantifies the levers available to management. A dashboard tells you revenue fell 12 percent. A briefing tells you the fall was churn-led, names the contribution of each component, states how much of next quarter is already contracted, and quantifies what can be recovered.
At minimum: an executive summary of the ranked material findings, and then a consistent treatment of revenue, gross margin, operating expenses, EBITDA, and cash and runway. Each area should answer the same four questions in the same order so that management learns the structure and can navigate it without guidance. External context such as central bank rates, inflation, and industry benchmark ranges helps management judge whether performance is company-specific or market-wide.
In a fractional CFO engagement, the CFO owns the briefing but should not be assembling it by hand each month. The assembly work — pulling figures, building bridges, calculating variances, formatting — is mechanical and should be automated from the close data. The CFO's contribution is judgment: which findings matter most for this business right now, what the recommended action is, and how to frame it for this particular management team.
Fynease Intelligence produces the briefing from your close data. Five operating lanes, four questions each, every figure traceable to the ledger.
See Fynease Intelligence Start free trial