A CFO's judgement, priced to the problem rather than to a headcount line.
It produced a number for the deck
and has not been opened since.
It produced a number for the deck and has not been opened since. A model that cannot reproduce last quarter is not a forecast, it is decoration — and when the board asks what happens if churn doubles, a driver-based model answers in one cell.
Neither tells you what to charge, which customers lose you money, or whether
the next hire is affordable.
Bookkeeping records what happened. Tax compliance files it.
Neither tells you what to charge, which customers lose you money, or whether the next hire is affordable. That is a third function entirely, and most companies discover the gap at the worst possible moment.
Runway moves when collections slip —
a 30-day customer becoming a 60-day customer can change the answer.
Most runway calculations divide cash by average burn. That is a number about spending, not about survival. Runway is a collections question: it moves when a large customer goes from thirty days to sixty, and founders get caught when they never model the gap.
Investors ask why gross margin moved four hundred basis points in Q3, and whether the cause is mix or pricing.
Investors and buyers spend little time on your revenue forecast.
They ask why gross margin moved four hundred basis points in Q3, and whether the cause is mix, pricing or an error. That question is either answered in advance or improvised in the room.
Twenty slides pulled from three systems, and the meeting still starts by debating the numbers.
Twenty slides assembled the night before from three systems, and the meeting is still spent arguing about whether the numbers are right instead of what to do about them. That is an architecture problem, not an effort problem, and no amount of working later fixes it.
Liquidation preference, participation, option pools, and SAFE conversion determine what you actually keep in
the outcomes that matter most.
Liquidation preference, participation, option pool top-up, conversion on the last SAFE. These decide what you keep in the outcomes you are most likely to get — and they are agreed in a week, by a founder seeing the structure once and an investor who has seen it 200 times.
The honest answer is not "as early as possible." It is when the cost of a wrong financial decision exceeds the cost of the engagement.
The most common trigger. A fractional CFO for startups earns their fee twice when the financial model is intact.The most common trigger. A fractional CFO for startups earns their fee twice over if they prevent one avoidable term-sheet concession — and the financial model is where those concessions usually originate.
Below that, a competent founder with a good bookkeeper can usually hold it together. Above it, the number of interacting decisions — pricing, hiring, working capital, credit terms.Below that, a competent founder with a good bookkeeper can usually hold it together. Above it, the number of interacting decisions — pricing, hiring, working capital, credit terms — exceeds what anyone can carry alongside running the company.
Lenders underwrite cash flow rather than growth, and covenant compliance is a monthly discipline rather than an annual.Lenders underwrite cash flow rather than growth, and covenant compliance is a monthly discipline rather than an annual one. A breach you did not see coming is materially worse than one you negotiated in advance.
Cohort economics, contribution margin by segment, a real downside case. If the ask has moved beyond a P&L, the reporting function needs to move with it.
No open-ended retainers that drift. Every engagement starts with the same diagnostic and then narrows to what you actually need.
A full read of what exists: historical accounts, any current model, the reporting pack, the cash position and the commitments against it.A full read of what exists: historical accounts, any current model, the reporting pack, the cash position and the commitments against it.
The output is a written assessment of what is sound, what is broken, and what it will cost to fix — with a recommendation to proceed or not.
Whatever the diagnostic identified as load-bearing. Usually a rebuilt three-statement model on real drivers, a 13-week cash flow forecast, and a reporting pack.Whatever the diagnostic identified as load-bearing. Usually a rebuilt three-statement model on real drivers, a 13-week cash flow forecast,
and a reporting pack that assembles in an hour rather than a week.
A monthly cycle: close review, variance against plan with the drivers explained, forecast update, and the two or three decisions that need a view this month.A monthly cycle: close review, variance against plan with the drivers explained, forecast update, and the two or three decisions that need
a view this month. Plus availability between cycles, because the questions that matter rarely wait for the monthly meeting.
Diligence-ready financials, valuation work, scenario and sensitivity analysis, and preparing you for the questions before they are asked — so that you can answer them yourself with the numbers behind you.Diligence-ready financials, valuation work, scenario and sensitivity analysis, and preparing you for the questions before they are asked
— so that you can answer them yourself with the numbers behind you. This is where the private-markets background does the most work.
Most founders arrive at a raise with a model built to produce a number for the deck, historical accounts nobody has stress-tested,
and no idea which three questions will actually be asked. The round then gets run on the investor's terms, against the investor's model.
ControlFi does not introduce investors, contact or pitch them, or negotiate terms on your behalf, and takes no fee contingent on a round closing.
Built on drivers, reproducing your historicals before it forecasts anything, with a balance check.
Updated weekly, showing the specific week cash gets tight under each scenario.
P&L, cash, KPIs, variance against plan with drivers explained, and a forward view.
Base, upside, and downside scenarios that move the underlying drivers—collections, churn, and conversion.
Contribution margin by segment, payback, and the
cohort view underneath them.
Diligence-ready financials, valuation support, data-room financial content, and the question list you will be asked.
Pricing is often opaque because three common engagement structures are compared as if they were equivalent.
Ongoing finance leadership. Fixed monthly fee, defined scope.
Fixed price for a model, reporting rebuild, or raise-readiness package.
Best for diagnostics and focused, time-bound engagements.
30+ cross-border M&A transactions and financings with $700M+ in
total aggregate value.Sell-side advisory on 30+ cross-border transactions and private financings worth $700M+, buy-side investing
across a $500M+ private markets portfolio, and in-house finance leadership — spanning EMEA, North America and APAC.
You work directly with the person doing the work. No junior handoff.No associate builds your model. No handover.
The diagnostic is designed to tell
you whether you need this, including
when the answer is no.The diagnostic is designed to tell you whether you need this, including when the answer is no.
ControlFi prepares you to raise —
the model, the financials, and the answers to what diligence will ask.ControlFi prepares you to raise — the model, the financials, and the answers to what diligence will ask.
We do not introduce investors, solicit them, or negotiate terms on your behalf, and it takes no fee contingent on a round closing.
Practical guides to hiring finance leadership, building reliable models, and understanding the numbers behind your business.
A 30-minute review to understand where you are, what is missing, and what should be next.A 30-minute review to understand where you are, what is missing, and what should happen next.
