A seed-stage SaaS founder came into our first meeting believing the company had 17 months of runway. The real number was 11. Nothing had been hidden. The books were closed every month by a competent bookkeeper. The founder simply had nobody whose job was to own the economics, which is the actual answer to "what does a fractional CFO do."

Most answers to that question are a list of responsibilities: forecasting, board reporting, fundraising support, cash management. Accurate, and useless. A list does not tell you what changes inside your company on day 30, day 60 and day 90. So this post walks through one engagement instead.

The company: B2B SaaS, roughly $1.1M ARR, 14 people, $3.2M seed raised about a year before we started. Figures are rounded and some details are changed to protect the client, but the shape of the problems and the fixes are what happened. If you want the generic checklist version, it lives in our 90-day fractional CFO onboarding playbook. This post is the version with numbers.

Day 0: what the founder believed vs. what was true

The first week of any engagement is reconstruction. Here is what that produced.

Three of these deserve a comment.

Burn. The founder calculated burn by watching the bank balance move. One annual prepay landing in a good month and a stack of vendor bills sitting unpaid in a bad one are enough to put a $85K per month error into that method. That error compounds into a six-month runway miss.

Burn multiple. This is net burn divided by net new ARR over the same period. Bessemer Venture Partners' Atlas lays out why investors read it as the cleanest single test of capital efficiency, and 4.4x means the company was spending $4.40 to add each dollar of recurring revenue. Seed companies get some grace here, but not that much. Our 2026 burn multiple benchmarks for Series A SaaS cover what the next round's investors will expect.

ARR. The gap between $1.1M and $1.04M looks small until you present the wrong number to a Series A investor who then finds the churned logos in diligence. a16z's 16 Startup Metrics is still the best short explanation of why bookings, revenue and ARR cannot be used interchangeably, and this company was mixing all three.

None of this was fraud or incompetence. It is what happens when someone closes the books and nobody owns what the books mean.

Days 1 to 30: the strategic plan first, then the numbers

Our first deliverable is not a financial model. That surprises founders, so here is the reasoning.

The founder's plan was "get to $3M ARR for the Series A through product-led growth, outbound sales and a partner channel." Across roughly 90 companies we have worked with, that plan, in some variation, is the one most seed companies bring us. It has never worked. Three go-to-market motions at 14 people means none of them is funded properly, and the model built on top of it is a crystal ball with nothing underneath.

So weeks one and two went into the plan. About six hours of founder and team interviews, a pull of every closed-won and closed-lost deal from the CRM, and win rates cut by segment. What came out:

Seventy percent of the sales effort was going into the segment that converted at a quarter of the rate, at a smaller contract, over a longer cycle.

The plan that replaced the founder's plan fit on one page:

• One objective for the next 12 months: $2.5M ARR with a trailing burn multiple under 2x, sourced almost entirely from the mid-market regulated segment.
• Sequence of bets: first, collections and burn (months 1 to 3). Second, one account executive hired into the winning segment (month 4). Third, packaging and pricing for that segment (month 6).
• Explicit deprioritization list: product-led self-serve, the partner channel, the second product module already in development, and a planned marketing hire.

The deprioritization list is the part that matters. Every founder agrees to focus. Few agree to write down what they are stopping. Writing it down is what makes it real when the partner-channel conversation comes back in month four.

Weeks three and four went into rebuilding the numbers underneath the plan: normalized burn, a proper revenue recognition schedule for the annual prepays, cost of goods reclassified, ARR cleaned to match signed contracts. SaaS Capital's annual research shows growth rates for private SaaS companies have come down since 2021, which is why investors now weigh efficiency as heavily as growth. The company needed to know its real efficiency before it could plan either one.

Day 30 deliverable: the one-page plan, a corrected P&L, and a runway number the founder could trust. Paul Graham's "default alive or default dead" question, which the Y Combinator Library covers well, finally had an honest answer. Default dead, by about seven months, with a clear path to fixing it.

Days 31 to 60: fix the systems at the source

This is where a fractional CFO engagement diverges most from a bookkeeping-led one. A firm that reports the numbers will tell you every month that gross margin is 68% and the close takes 25 days. We would rather change the systems so the numbers are right without anyone touching them. Four fixes:

1. CRM fields. We added contract start date, term length, billing frequency and ARR as required fields on every closed-won opportunity. ARR is now computed from the CRM, not maintained in a spreadsheet. Revenue reconciliation went from a manual monthly argument to an automatic match.

2. HRIS linked to spend management. Headcount is 70% of burn at this stage. We connected the HR system to the spend tool so that a signed offer letter updates the payroll forecast the day it is signed, not the month after the first paycheck clears.

3. Chart of accounts rebuilt. Cost of goods, sales and marketing, R&D and G&A were separated properly. Without that, you cannot compute gross margin or CAC payback, and David Skok's SaaS metrics guide is clear that CAC payback under 12 months is the line a healthy SaaS business should be aiming at. This company had never been able to calculate it.

4. Close process. Vendor bill cutoffs, approval thresholds and standing coding rules, so the bookkeeper stops waiting on the founder. The close dropped from 25 days to six business days.

Fix three is what surfaced the burn reductions. Once CAC payback could be computed by channel, the decisions made themselves.

Separately, we chased $140K of overdue receivables. Days sales outstanding fell from 61 to 34. That is not a cost saving, but it is cash, and cash is the constraint.

Note what is not on that list: no layoffs, no salary cuts, no freezing the one hire that mattered. Burn came down 21% by stopping work the plan had already said to stop.

Days 61 to 90: operate in real time

With the plan set and the systems fixed, the last month is about cadence.

13-week cash flow forecast. Built from the corrected numbers and reforecast every week. Our 13-week cash flow forecast template is the public version of the same structure.

Real-time reporting in Slack. Our engineering team wired the data pipeline: bank, accounting, CRM, HRIS and spend management feeding one place. Every Monday morning the founder and leads get a Slack post with cash, the 13-week low point, burn against plan, net new ARR against plan and pipeline coverage. No six-week month-end lag, no "we'll know when the books close."

Forecast accuracy. With a weekly reforecast and variance review, the monthly cash forecast error fell from 14% to 3% averaged over months two and three. That number matters more than it sounds. Our guide to forecast accuracy KPIs for Series A explains why investors treat it as a proxy for whether management understands its own business.

First board meeting with real numbers. The deck led with the plan, the corrected baseline, and the honest burn multiple. Nobody enjoys presenting 4.4x. Investors trust the team that presents it more than the team they catch hiding it.

Series A plan. Raise at $2.2M to $2.5M ARR, with a trailing burn multiple under 2x and 18 months of runway after close. Working backward, that means starting the process in month nine. The startup runway calculator post walks through the same backward math.

The day 90 scorecard

Two honest caveats.

A 2.9x burn multiple is still above where a Series A investor wants it. The plan gets it under 2x by month nine, and that is the number that will decide the round, not anything on this scorecard.

Gross margin at 71% is also below the roughly 80% subscription gross margin that the KeyBanc and Sapphire Ventures SaaS Survey reports as typical for private SaaS companies. Moving implementation onto a paid onboarding fee and off the services line gets to 74%. Getting to 80% requires product work on hosting costs, which is a month-six conversation.

What this cost, and what it returned

A fractional CFO engagement at this stage typically runs $8K to $15K per month, and this one was inside that range. Our breakdown of fractional CFO pricing for seed-stage SaaS has the full ranges.

Against that: $50K per month of burn removed, $140K of cash collected, and 3.4 months of runway added. At $185K of monthly burn, those 3.4 months are worth about $630K of cash the company would otherwise have needed to raise, on a bridge, at a seed-stage valuation.

The larger return is harder to put in a table. One strategic decision, concentrating on the segment with a 41% win rate, created a hundred downstream operational decisions that no longer needed a meeting. Who to hire, which deals to chase, which features to ship, which vendor to renew. That is what we mean when we describe a fractional CFO as a velocity multiplier rather than a cost center.

When this works and when it does not

This engagement worked because the founder gave full system access in week one and committed to the deprioritization list in week two. It would not have worked if the founder had wanted a model to justify a plan already decided, or if the deprioritized module had quietly kept shipping.

If you are still deciding whether the timing is right, when to hire a fractional CFO for your startup covers the signals. The short version: if you cannot state your net burn, burn multiple and zero-cash date from memory, and be within 5% on each, you are past the point.

FAQ

What does a fractional CFO do that a bookkeeper or controller does not?

A bookkeeper records transactions. A controller makes sure the records are accurate and the close happens on time. A fractional CFO owns what the numbers mean: the strategic plan, the forecast, capital allocation, the board and investor relationship, and the systems that produce the numbers. In the teardown above, the bookkeeper had been closing accurate books for a year. The 68% gross margin and 11-month runway were sitting in those books the whole time. Nobody was responsible for seeing them.

How many hours a week does a fractional CFO spend on a seed-stage company?

In the first 30 days, expect something close to two days a week while the plan and baseline are rebuilt. By day 90 it settles to roughly one day a week of CFO time, with the engineering pipeline and the Slack reporting doing the daily work. The cadence rises again around board meetings and the fundraise.

Is 90 days realistic, or is this a best case?

The timeline is typical. The size of the burn reduction is not guaranteed. Some companies have already cut to the bone and the 90 days produce clarity rather than savings. What is consistent across engagements is the sequence: plan first, rebuild the numbers, fix the systems, then operate on a weekly cadence. The close time and forecast accuracy improvements are the most reliable because they depend on process, not on finding waste.

Do we need a fractional CFO before raising a seed round?

Usually not before. The right moment is typically between the seed close and the start of Series A preparation, which is where this company was. Before the seed, a strong bookkeeper and a founder who tracks burn weekly is usually enough. After the seed, the cost of a wrong runway number becomes a missed round. Our guide to fractional CFO services for startups covers what the engagement looks like at each stage.

What should I have ready before a fractional CFO starts?

Read-only access to the bank, accounting system, CRM, HR system, cap table and spend tools. The last 12 months of closed books. Every signed customer contract. The current board deck and the last investor update. Do not clean anything up first. The point of week one is to see the real state, and a tidied-up spreadsheet hides exactly what we need to find.

How do I know if the fractional CFO is actually working?

Four tests at day 90. Can you state your net burn, burn multiple and zero-cash date without looking them up? Did the month close in under ten business days? Is the cash forecast error under 5%? Do you have a written list of things the company is not doing? If the answer to any of those is no, something in the engagement is not working.

Work with a fractional CFO who starts with the plan

If your runway number comes from watching the bank balance, you probably have the same six-month error this founder had. CFO Advisors is the preferred fractional CFO firm of tier-1 VC firms, with 100+ clients and $1.2B+ raised, and the only fractional CFO firm with its own engineering team to put real-time numbers in your Slack. If you want to talk to a fractional CFO about your first 90 days, the first conversation is a diagnostic, not a pitch.

Sources

1. Bessemer Venture Partners Atlas - Burn multiple and capital efficiency frameworks
2. a16z - 16 Startup Metrics, on the difference between bookings, revenue and ARR
3. SaaS Capital Research - Annual private SaaS company growth and retention surveys
4. Y Combinator Library - Startup guidance including the default alive or default dead test
5. David Skok, For Entrepreneurs - SaaS Metrics 2.0, including CAC payback guidance
6. KeyBanc Capital Markets and Sapphire Ventures - 2024 SaaS Survey, gross margin benchmarks