Fractional CFO for Professional Services Firms

In a product business, revenue problems show up in inventory and orders. In a services firm they hide in softer places: a utilization percentage nobody computes the same way twice, projects that were profitable in the proposal and break-even in delivery, and a pipeline everyone describes with adjectives instead of numbers. Meanwhile the cost side is brutally simple: salaries, every two weeks, regardless.

The CFO engagement instruments the whole machine: a revenue forecast built from headcount, utilization, and realized rates; project-level profitability that shows which work and which clients actually make money; pricing models for hourly, fixed-fee, and retainer structures; hiring math tied to pipeline coverage; and the partner-level economics (comp, promotion, buy-in) that most firms improvise. It is delivered as a monthly cadence on top of clean books, not a one-time strategy deck.

For the general engagement model, scope, cadence, and how fractional delivery works, see the fractional CFO services page. For the accounting foundation specific to consultancies, our accounting for consultants page covers the bookkeeping and project-accounting layer this builds on.

The engagement usually earns its fee when several of these are true:

• The firm has roughly 10 to 100 people, and payroll is 50% or more of revenue.
• Revenue crossed a few million dollars, but profit has been flat for two or more years while headcount grew.
• Nobody can produce project-level margins that the delivery leads actually believe.
• Rates have not changed in two or more years, or change only when a client pushes back less than expected.
• Hiring happens ahead of signed work on instinct, and the bench after a project ends is absorbed in silence.
• One client is a large share of revenue, and their payment terms effectively set the firm's cash policy.
• A promotion to partner is pending and nobody has priced what it costs or requires.

If only the bookkeeping is shaky, fix that first; the CFO layer multiplies a good foundation and flatters a bad one. A free initial consultation sorts out which layer your firm actually needs next.

Most services firms forecast revenue as last year plus a growth percentage. The utilization-driven model replaces that with the actual engine:

Supply side: what the team can deliver. Billable headcount times capacity hours, times a utilization assumption set per role and per person from history, times the realized rate (the rate actually collected after write-downs, not the rate card). This is the revenue ceiling, and it moves only when headcount, utilization, or rates move.

Demand side: what the pipeline can feed. Signed backlog scheduled over the months it will actually be delivered, plus the weighted pipeline: each proposal times a probability the partners defend with a straight face, reviewed monthly. The ratio of weighted pipeline to the revenue gap is the firm's pipeline coverage, and it is the number that should gate every hire.

The forecast is the reconciliation of the two. Supply above demand means bench cost and a selling problem. Demand above supply means turned-away work, burnout risk, and a hiring or pricing decision. The model shows which regime the firm is in for each of the next twelve months, which is precisely the thing instinct gets wrong at the turns.

1. Payroll is fixed while revenue is project-shaped. Projects end on Fridays; salaries continue on Mondays. A firm that staffs to its best quarter carries the bench through its worst, and the cash gap lands exactly when morale is already low. The forecast treats bench cost as a first-class number, sets a working-capital floor measured in payrolls (most firms should hold several), and gates hiring on pipeline coverage rather than on how busy everyone feels this month.

2. Scope creep and unbilled work: the margin leaks nobody logs. Fixed-fee projects absorb "one quick addition" at a time until the effective rate is embarrassing, and hourly projects leak through recorded-but-written-down time. Neither shows up on the P&L as a line item; both show up as a firm that works hard and earns average margins. Project-level profitability reporting, actual hours at cost against fees collected, makes the leak visible per project and per client, which is the prerequisite for fixing it with change orders and pricing.

3. Client concentration sets the firm's payment terms. When one client is 30% or more of revenue, their 75-day payment behavior becomes the firm's cash policy, unilaterally. The CFO tracks concentration and AR by client, prices payment terms into new engagements (deposits, milestone billing, shorter terms for smaller clients), and builds the cash floor assuming the anchor client's worst recent behavior, not their contract.

Pricing structure. The model prices the firm's actual work three ways against historical delivery data: hourly (simple, but caps upside and invites scrutiny of every entry), fixed fee (higher margins when scoped from real effort data, dangerous when scoped from optimism), and retainer (smooths cash beautifully, quietly loss-making when heavy clients are never re-priced). Most firms should run a mix, and the mix should be a decision, not an accident of which client asked for what.

Leverage. The ratio of partners to seniors to staff on delivery determines blended cost per delivered hour, and it is the least-examined number in most firms. Partner-heavy delivery earns premium rates at premium cost and cannot scale; leveraged delivery drops blended cost faster than blended rate and frees partner hours for selling and reviewing, which is where partner time actually earns its keep.

Firms model client projects to the dollar and then admit partners on sentiment. The partner-track model treats admission as what it is, a permanent claim on profits, and prices it:

The accretion test. What must the candidate originate and deliver, at what leverage, for existing partners' income to be higher with them as a partner than without? The model turns "they deserve it" into a number everyone can see.

Buy-in structure. How the equity is valued, funded (capital contribution, comp reduction over time, or a note), and what it purchases: profit share, vote, or both. Modeled in good-year and bad-year versions before terms are offered.

The comp formula under stress. Whatever split the firm uses (equal, tiered, origination-weighted), the model runs it against the last three actual years, including the bad one, so the partnership discovers how the formula behaves before it is binding.

Exit symmetry. The same math, run in reverse, prices retirements and departures, which is when unmodeled partnerships get expensive and personal.

Partner admissions also carry entity and tax consequences for the firm and the new partner, which we coordinate with the tax side before terms are final rather than discovering them at filing time.

• P&L against budget with labor cost split into billable and non-billable, and variances explained.
• Utilization report by person and team against role targets, from the time system, one agreed definition.
• Project profitability: effective rate and margin per active project, with fixed-fee burn tracked against percent complete.
• Client profitability and concentration: margin and AR by client, top-client share of revenue trended.
• Pipeline and backlog: signed backlog scheduled by month, weighted pipeline, and the coverage ratio that gates hiring.
• Headcount plan: current roster against the forecast's supply-demand picture, with any proposed hire's math attached.
• 13-week cash forecast with the working-capital floor and days-payroll-on-hand flagged.

The pack ends with a one-page decision list: what needs a call this month, what the numbers say, and what the CFO recommends. Reporting that does not end in decisions is scrapbooking.

Definitions live in one place and the numbers come from the time and accounting systems automatically. The full definitions, with formulas, are on the service business KPI dashboard page.

Monthly. The pack is reviewed with leadership: utilization, project margins, pipeline coverage, cash. Underwater projects get a recovery plan or a change order; hiring requests get tested against coverage on the spot.

Quarterly. Pricing pulse (realized rates vs rate card by client), client portfolio review (who gets invested in, who gets re-priced, who gets gracefully released), comp accrual true-ups, and coordination with the tax team on estimated payments and owner comp from the same forecast.

Annually. The big three: the rate letter (priced from realization data, not bravery), the budget and headcount plan by service line, and the partner-track review, promotions, comp formula, and any admission or exit modeled before it is discussed with candidates.

• Revenue grew and profit did not, for the second year running, and the partners are out of theories.
• A hiring wave is proposed and someone finally asked what the pipeline actually supports.
• A partner decision is pending: a promotion, an admission, a retirement, or a departure, with real money attached and no model.
• A big client got bigger, and with it the quiet realization that the firm's cash flow now belongs to someone else's AP department.
• An acquisition conversation started, inbound or outbound, and the firm needs numbers a buyer's diligence team will respect.

For what the engagement should cost, the fractional CFO cost guide covers the drivers; for the in-house alternative, the fractional vs full-time CFO comparison covers the staffing decision head-on.

• Time tracking that is actually complete, billable and non-billable, entered weekly, because utilization computed from partial data flatters everyone and informs no one.
• Project-level accounting: revenue and labor cost mapped to projects and clients in the ledger, so profitability is a report, not a quarterly research project.
• A monthly close on a schedule, with revenue recognized consistently for fixed-fee and milestone work rather than when the invoice happens to go out.
• Payroll mapped to roles and service lines, the input that leverage and pricing models run on.
• AR discipline: invoices out on schedule, aging worked weekly, and deposits or milestone terms on every new engagement.

When this layer is missing, we scope the cleanup first through the bookkeeping side (see accounting for consultants), then start the CFO cadence. A utilization model on top of half-entered timesheets is a very confident way to be wrong.