What does working together actually look like?
Most pages about fractional CFO services describe a category. This one describes the shape of the work: how an engagement is scoped, what happens in a month, what you receive, and what it needs from you to be worth anything.
It starts with what is happening, not with a service
Founders rarely arrive wanting a service. They arrive with a situation: an investor asking for something they cannot produce, a margin that is falling for no visible reason, a board meeting in three weeks, a decision about a location or a product line that nobody can put numbers behind.
So the first conversation is about that, not about scope. What is actually breaking, how long it has been breaking, what has already been tried, and what would have to be true for it to stop mattering. Twenty minutes of that usually makes the shape of the work obvious, and it occasionally makes it obvious that there is no work to do, which is a legitimate outcome.
What follows is a description of the shapes an engagement takes once that conversation has happened.
Three shapes, depending on what is wrong
- A question answered
- A specific decision that needs analysis and then stops needing it. Should we take this customer on these terms. What does the new product line have to sell to pay for itself. Can we afford this hiring plan. Scoped to the question, delivered as the work and the reasoning, and finished when the decision is made.
- A defined piece of work
- Something that has a beginning and an end and leaves an artefact behind. A financial model for a raise. A reporting pack built and handed over. Diligence preparation. An annual operating plan. Scoped to the deliverable, with the handover built in so it keeps working after the engagement ends.
- An ongoing rhythm
- A recurring monthly commitment: the pack, the cash view, the questions that come up, and the two or three decisions a quarter that are worth real analysis. This is the shape that suits a company whose problem is not one thing but the absence of anyone whose job is to look.
Most companies start with the first or the second. Some of those become the third and many do not, which is the point of starting small: a defined piece of work is a cheap way to find out whether the working relationship is any good before either side commits to a rhythm.
A month, when it is ongoing
The rhythm matters more than the hours. A month that happens on a schedule is worth considerably more than the same amount of attention delivered when someone remembers.
- Early in the month: work with your accountant to get the books closed on a date rather than eventually, and check what came out of them before anyone builds on it.
- Then the pack: the same format every month, with the numbers against plan and a written explanation of what moved and why. Written to be read by someone who was not in the room.
- The cash view, updated: what is committed, what is likely, and where the tight weeks are over the next thirteen.
- A review conversation: what the numbers say, what they do not say, and what decision is now in front of you.
- Whatever the quarter actually needs: a pricing question, a hiring plan, a board pack, an investor request, a model that has stopped matching reality.
- And availability between all of that, because the useful questions arrive on Tuesday rather than at month end.
Nothing in that list is unusual. What makes it work is that it happens on the same dates every month, so decisions stop waiting for numbers.
What gets handed over
Deliverables described as what you receive rather than as what they will achieve, because the second is not something anyone can promise on a web page.
- A monthly reporting pack in a fixed format: performance against plan, cash and runway, the operating metrics that matter for your business, and written commentary.
- A cash forecast you can update yourself, covering the next thirteen weeks, with the statutory calendar in it.
- A financial model built on drivers rather than a growth rate, with the assumptions visible and separated from the arithmetic.
- An annual operating plan for an April to March year, with the targets, the headcount plan at fully loaded cost, and the cash consequence.
- Unit economics by whatever segment actually differs in your business, with the allocation rules written down so the analysis can be reproduced next quarter.
- Analysis for specific decisions, with the working shown rather than a conclusion.
- And the definitions: what each metric means, written once, so the numbers still mean the same thing in a year.
Everything is built to be handed over. A model only its author can update, or a pack that stops being produced when the engagement ends, has not solved the problem it was brought in for.
What it depends on from your side
This is the part most engagement descriptions leave out, and it decides whether the work is any good.
- 01Books that get closed. Not perfect books, and not fast ones to begin with, but an accountant who closes the month and someone who can answer questions about what is in them. Analysis on numbers nobody has checked is worse than no analysis.
- 02Access to the actual data: the accounting system, the billing or POS system, the bank, and whatever spreadsheet the business really runs on.
- 03One person inside the company who owns the relationship and can get answers. It does not have to be the founder, and it cannot be nobody.
- 04Honesty about what is wrong. Engagements go badly when the real problem is not the stated one, and that usually surfaces in month three rather than week one.
- 05A decision you are actually willing to make. Analysis that nobody intends to act on is an expensive way to feel informed.
How it starts
- 01A conversation about what is happening. No deck, no prepared material, and no obligation. If there is an obvious answer that does not involve hiring anyone, you will get it here.
- 02If it is worth going further: a look at what actually exists. The last few months of accounts, whatever reporting there is, and the spreadsheet the business runs on. This is where the real scope becomes visible, and it frequently differs from the stated one.
- 03A written proposal: what would be done, in what order, what is handed over, and what it depends on. Specific enough that you could give it to someone else to deliver.
- 04Then the work, starting with whatever is most broken rather than with whatever is most impressive.
Pricing is discussed in that conversation rather than published here, because it depends on the shape of the work and on what condition the numbers are in. What is published on this site is how to compare proposals properly, which is a more useful thing for a founder to have before any conversation.
Where engagements go wrong
- Scope that was never written down, so two people have different pictures of what is being delivered and both are disappointed in month two.
- The real problem being different from the stated one. A company that asks for a model and actually needs a close is a common version of this.
- No single owner inside the business, so every question waits a week and the work happens at a quarter of its useful speed.
- Analysis delivered into a vacuum, where nobody was ever going to act on it.
- An engagement that becomes reporting production rather than judgement. If the monthly pack is the whole of it after six months, something has gone wrong with the arrangement.
- And the opposite: advice with no artefact, where nothing is left behind that survives the engagement.
Most of these are visible early if anyone is looking for them, which is why the first review conversation should include whether the arrangement itself is working rather than only what the numbers say.
What this is not
- Not bookkeeping or accounting. Your CA firm keeps the books, files the returns and handles the audit, and that arrangement should continue.
- Not statutory compliance or tax advice. Those are specialist work and they stay with the people qualified to do them.
- Not audit, and not a substitute for one.
- Not fundraising. The model, the data room and the diligence preparation are finance work. Introductions and running a process are a different job done by different people.
- Not legal advice on term sheets or shareholders' agreements, though the arithmetic underneath them is fair game.
- Not a full-time CFO, and not a stand-in for one where a company genuinely needs that.
The boundary is worth being clear about, because an adviser who claims all of it is describing a category rather than a job. What sits in the middle here is the part most early-stage companies are missing: someone whose work is to look at the numbers and say what they mean, and then build whatever is needed so the next question is easier to answer than the last one.