A Fractional CFO First 90 Days: What Gets Fixed in Which Month and How You Measure It

By Andreas Pettersson, Founder, Leaders ADAPT

Ninety days is long enough to change how a company sees its money and short enough that nobody gets comfortable. That is why I like it as the first window for a fractional CFO. It is also why I ask owners to write down, before the engagement starts, what they expect to be different on day 90. Most cannot. They know the books feel wrong. They do not know what right would look like.

A fractional CFO first 90 days runs in three months with three jobs: diagnose, fix, run. Month one makes the numbers visible, month two makes them useful, month three makes them routine. Each month has its own deliverables and its own measures, and the owner should be able to check every one of them without being an accountant.

Why ninety days, and why in this order

I ran Arcules from zero to over 150 people, and for a stretch I held the CEO, CFO and CTO titles at once. The lesson from that stretch: when the numbers are late, every other decision waits for them, or worse, gets made without them. Sequence matters more than speed.

The order is diagnose, fix, run because each step depends on the one before. You cannot price with confidence until you trust the margin. You cannot trust the margin until the close is clean. You cannot build a forecast the bank believes until the cash view matches the bank account. A fractional CFO who starts with the forecast is decorating a house with no foundation.

McKinsey's survey of new CFOs found that more than 55 percent named understanding the business drivers as the most critical activity of their first hundred days, and about three quarters started or planned fundamental changes to core finance activities (McKinsey, 2008). The research is old and aimed at large companies. The pattern holds in a $15M company. Learn first, then change the machine.

If you are still deciding whether you need the seat at all, start with our fractional CFO guide, which covers what the role is and when to switch to a full time hire. This page assumes the decision is made and the person starts on Monday.

Month one: diagnose the close, the cash and the accounts

Month one is reading, not redesigning. The fractional CFO earns the right to change things by showing the owner what is actually there.

The close. How many business days after month end do the books close, who does what, and where does it stall? The CFO sits through one full close, watches it, and writes down every step that waits on a person. In many owner led companies the close is not a process. It is one bookkeeper and a long weekend.

Cash visibility. Can anyone in the company say what cash will be in six weeks? The Federal Reserve's Small Business Credit Survey found that 51 percent of employer firms reported uneven cash flows as a financial challenge and 56 percent reported trouble paying operating expenses (2024 survey, published March 2025). Uneven cash is normal. Being surprised by it is optional.

The 13 week cash flow. Built in the first two weeks, run every Monday after. The test is simple: forecast closing cash for the week, then compare it with the bank balance. The gap is the first number on the scorecard. JPMorgan Chase Institute found the median small business holds 27 cash buffer days, with a quarter holding fewer than 13 (2016). Your company may sit far above that. The forecast tells you whether you do, every week, before the bank does.

The chart of accounts. Map it, do not rebuild it yet. The CFO looks for revenue lumped into one line, owner costs mixed with business costs, and cost of goods hiding in overhead. Those are the places margin goes missing. The agreed new structure goes live at the start of the next quarter, so comparisons stay clean.

How the owner measures month one:

• A written diagnosis on day 30, three pages at most, in plain language.
• The 13 week cash flow has run for at least two Mondays.
• The close calendar is written down, with a name on every step.
• You can name the three biggest finance risks in the company without looking at notes.

If you get forty slides on day 30, push back. Month one produces a short read and a working cash forecast. Nothing else.

Month two: fix pricing, margin, reporting and the bank

Month two is where the owner starts to feel the seat. The CFO now picks the fixes that pay for the engagement.

Pricing and margin. Gross margin by customer, by product line, by location, whichever cut matters in your business. The first version is usually uncomfortable. A customer you are proud of turns out to be your thinnest margin. A product line you were about to drop turns out to carry the quarter. The CFO's job is to show you this with numbers you trust, and then to recommend one pricing move with a figure and a date.

The reporting pack. One page, before the leadership meeting, every month. Cash now and in 13 weeks. Revenue and gross margin against plan. The three numbers that moved and why. The one decision the month needs. Appendices exist for the people who want them. The test is not whether the pack is accurate. It is whether the owner reads it before the meeting.

The KPIs. Five to eight finance numbers that sit on the weekly leadership scorecard, each with an owner who is not the CFO. Days sales outstanding belongs to whoever runs collections. Gross margin by line belongs to whoever sets price. If your team runs a weekly scorecard, the CFO adds the finance lines to it rather than building a second one.

The lender relationship. Now, not in week one. Once the close ties out and the forecast is holding, the CFO meets the relationship manager with a short pack: the covenant position, the forecast and the one question the bank has been waiting to ask. Arriving with numbers you can defend changes the tone of every later conversation.

How the owner measures month two:

• The second close is faster than the first, and the date is published.
• Gross margin by line exists, and you have acted on at least one finding.
• The one page pack went out before the leadership meeting, and you read it.
• The bank has met the CFO and has a forecast in hand.

Month three: run the cadence, the forecast and the handover

Month three tests whether any of it sticks when the CFO is not in the building.

The monthly cadence. Close by a fixed day. Pack out by the next fixed day. A finance slot in the monthly leadership review with the same agenda every time. The CFO runs the first one and watches the team run the second. Ninety days in, the cadence should belong to the company, not to the CFO's calendar.

The forecast. A twelve month forecast built on the clean chart of accounts and the margin work from month two, with two or three scenarios the owner can argue with. Not a budget exercise. A tool for the decisions coming in the next two quarters: a hire, a price change, a line of credit.

The owner and board numbers conversation. If you have a board, an advisory board or a set of partners, month three is the first meeting where the CFO presents and the owner decides. If you have none, the conversation is between you and the CFO, once a month, with the pack on the table and the phone face down. Either way, the measure is whether a decision got made with a number attached.

The handover plan. A written plan for months four to six, and a written view of the end state. Does this company need the fractional seat for a year, a lighter CFO advisory shape after the cadence runs, or a full time hire the CFO helps recruit? A fractional CFO who cannot answer that at day 90 has not finished the diagnosis.

How the owner measures month three:

• The team ran one monthly review without the CFO leading it.
• The twelve month forecast exists, and you used it in one real decision.
• A written plan for months four to six sits in your inbox.
• You know which shape of the seat comes next.

The 90 day scorecard

Five lines. Write the starting value in week one and the target for day 90 in the engagement letter. Review at day 30, day 60 and day 90. The targets below are the ones I start the conversation with in a company between $5M and $50M. Yours depend on where you begin.

The last line matters most and gets measured least. Count the decisions in your leadership meeting that came with a number, a risk and a date. In week one the honest count is often zero. On day 90 it should be the normal way the meeting works.

Our page on how fractional engagements work covers the contract side of the same window: notice, the review clause and who owns the work. Market prices for the seat sit in the sourced fractional CFO cost table.

Where AI fits in the first 90 days

Gartner's 2025 survey of 183 finance leaders found that 59 percent use AI in their finance function, with knowledge management, accounts payable automation and error and anomaly detection as the top uses (Gartner, November 2025). Those three map well onto the 90 days.

In month one, anomaly detection speeds the diagnosis. A finance lead with a modern ledger and an assistant can scan a year of transactions for miscoded entries and duplicate payments in an afternoon, then spend the saved hours with the people who made them.

In month two, AI drafts the commentary under the pack. The CFO edits it. The owner gets a page that explains the movement in plain words without waiting a week for someone to write it.

In month three, knowledge management keeps the cadence alive. The close checklist, the definitions behind every KPI and the reasons for each forecast assumption go into one searchable place, so the next person does not start from zero. In an EOS style company this lands in the quarterly rhythm, which our AI for EOS companies guide covers.

Where AI does not fit: the pricing call, the conversation with the bank, the decision on what the company does next. Our page on what not to delegate to AI is the long answer. The short one: a model can show the CFO a margin problem. Only you can decide to raise the price to the customer you have had for twelve years.

What the first 90 days does not fix

It does not fix an owner who is the only person allowed to decide. A clean close and a good pack expose that dependence faster than anything else. Every number in the pack points back to one desk. Changing it is leadership work, which is why I pair the CFO seat with leadership team development when the scorecard shows decisions still queuing at the owner.

It does not fix a missing bookkeeper or controller. If the month one diagnosis says the ledger itself is broken, the honest recommendation is to hire the person who keeps it, then bring the CFO back to read it. Paying CFO rates for data entry is a mistake I will tell you about on the first call.

It does not fix a revenue problem. If the margin work shows the issue is pipeline, not price, the next seat is a fractional CRO, and the CFO's job is to say so by day 60.

It does not fix everything in 90 days. Ninety days sets up the system. The value shows up in the decisions you make in months four to twelve with numbers you trust.

Who is on the bench for this

Steve Shaw is the fractional CFO on the Leaders ADAPT bench. Steve has twenty years in finance leadership and works with owners on clean books, cash discipline, financial positioning before a sale, and the numbers conversation with buyers and lenders. In a first 90 days that means he runs the diagnosis himself, builds the cash forecast with your team, and sits next to you in the first meeting with the bank. Where the operating problem is finance led, Steve also takes COO seats covering cash, margin, pricing and the cadence around them.

Sergey Dmitriev is the fractional CTO and AI systems partner on the bench. When the reporting pack needs a data stack underneath it, or the close checklist and KPI definitions need a company knowledge system so they outlive the engagement, Sergey builds and runs it, and trains the leadership team on the tools with me.

Harvard Business Review describes fractional leaders as people who typically have 20 to 30 years of experience (HBR, July 2024). Experience is the entry ticket. What you are buying in the first 90 days is someone who has seen the sequence go wrong before and does not repeat it.

Every engagement runs through Leaders ADAPT. I take the first call, match the partner, agree the scorecard with you, and stay in the room through day 90 and after. The full bench, across finance, operations, technology and revenue, is on the fractional executive services page, and the product page for joining is Fractional Executives and Advisory.

Fractional CFO First 90 Days FAQ

What should a fractional CFO deliver in the first 90 days?

By day 30, a written diagnosis of the close, cash and chart of accounts, plus a 13 week cash flow running every week. By day 60, a one page reporting pack, gross margin by customer or product, and a reset lender relationship. By day 90, a monthly cadence the team runs, a twelve month forecast, and a written plan for months four to six.

How fast should the month end close get after a fractional CFO starts?

Expect the first close to be slow, because the CFO is reading every account. The second should be faster, and by the third month a company between $5M and $50M should close inside ten to fifteen business days, with a published calendar. If the date has not moved by day 90, the cause is usually a people or system gap the CFO should have named.

How do I measure whether a fractional CFO is working?

Track five numbers from day one. Days to close, the gap between forecast cash and the bank balance, gross margin by line, the share of reports read before the meeting, and the number of decisions made with a number attached. Write the day 90 target for each in the engagement letter. Then review them together at day 30, day 60 and day 90.

Should a fractional CFO change the chart of accounts in month one?

Map it in month one, change it at the start of a quarter. Rebuilding the chart mid quarter breaks every comparison the owner and the lender rely on. The month one job is to find the accounts that hide margin, mix owner costs with business costs, or lump revenue lines together, and to agree the new structure so it goes live on a clean date.

When does a fractional CFO talk to my bank?

Usually in month two, once the numbers are trusted. Calling the lender in week one with books that do not tie out costs credibility you then spend months rebuilding. The better sequence is to fix the close and the cash forecast first, then meet the relationship manager with a short pack, the covenant position and a forecast the CFO is willing to defend.

What if the first 90 days do not go well?

Say so at the day 90 review, plainly. Look at the scorecard, not at effort. If the close is no faster, cash still surprises you and nobody uses the pack, either the scope was wrong, the person is wrong, or the company needed a controller first. A good fractional CFO names which one before you have to ask.

How many days a week does a fractional CFO need in the first 90 days?

Often more than later. A company with messy books may need two days a week through month one and month two, dropping to one day once the cadence runs. Agree the front loading in advance and write it into the scope, so the extra days in the first quarter are planned, not discovered on an invoice.

Put day 90 in the calendar today

Owners tend to judge a new finance leader on how busy they look. Busy is not the measure. The measure is five lines on a scorecard and a meeting on day 90 where you read them together.

Open your calendar and book that meeting before the engagement starts, ninety days from the start date, two hours, with the scorecard as the only agenda. It changes the first 89 days, because everyone knows the date.