Customer Profitability Analysis: Formula, Steps, Forecast

Customer profitability analysis (CPA) is a client-level accounting practice that ties revenue to the cost of delivering the work. It measures what each account brings in against what it costs to serve. One company-wide figure hides which relationships are subsidizing the rest. Services firms call it client profitability analysis. Product profitability analysis applies the same logic to a product line.

The cost side is built from timesheets, and timesheets move every week. Two accounts on identical fees can land far apart. The gap comes from who does the work and how many hours get logged. Unbilled scope the team absorbs widens it further.

A senior-heavy delivery mix costs more per hour than a junior one. That is one of the key drivers of agency profitability.

The arithmetic runs like this. Billable hours at cost rate, plus non-billable account management and rework, plus an overhead allocation. Set that total against invoiced revenue for the same period.

You calculate customer profitability by subtracting the cost of serving a client from what you invoiced them. Divide what is left by that revenue to get a percentage.

Pull the invoice total for the period. Price the logged hours at each person’s fully loaded cost rate (salary, benefits, and payroll taxes). Add subcontractor fees and a share of overhead. Which costs you subtract depends on what you sell.

Take a brand agency client billed $12,000 for a website build.

Both figures belong in your agency profitability metrics.

At 43% gross on a fixed fee there is not much room left. Model the same job at 110 hours. Delivery cost rises to $9,350 and gross margin falls to 22%.

Read that figure next to utilization and the other agency KPIs. It tells you whether the cause is your price or your hours.

A spreadsheet handles this fine for five clients. Past that, hours, cost rates and invoices sit in separate files that someone refreshes weekly. Every figure you quote is already out of date.

Budgeting in Productive holds the rate card per client and tracks time and costs as work runs.

Two formulas do the work. Delivery costs are logged hours at fully loaded cost rates, plus subcontractor and expense costs.

With your profitability results, sort every client from highest to lowest margin. Then work the two ends of that list. At the bottom sit the accounts earning below your target. Reprice them, change who delivers the work, or drop them.

At the top sit the accounts that already fund the business. Those get protected staffing and the next upsell conversation.

Understanding customer profitability gives you a margin ranking, better-informed marketing spend, and sharper service line decisions. It also shows where costs come from, strengthens retention, and produces a year-end number you can defend.

Each one changes a different decision, from who you pitch next to who you staff first.

The main steps are: collect data, assign costs, calculate margins, segment clients, plan actions, and review quarterly. Below, we break down the whole process in steps you can replicate.

Export three files covering the same 90 days. Invoiced revenue per client from billing, logged hours per person per client, and subcontractor invoices. Add each person’s cost rate as a column on the hours file.

You are done when no week of time is missing. Chase the gaps with the people who own them before you calculate anything. A spreadsheet someone refreshes by hand dates every figure you quote.

Price each person’s logged hours at their fully loaded cost rate. Split overhead across total hours worked so no account carries another’s rent. A $5,000 monthly facility cost across 488 worked hours adds $10.25 to every logged hour.

Calculating agency overhead covers what belongs in that pool. Apply the same per-hour figure to every account, including the small ones.

Subtract delivery costs and allocated overhead from invoiced revenue for the same period. Divide the result by revenue. Run it per client and per month. A 90-day average lets a bad March hide inside a good quarter. Set the flag line at your own target margin and mark every account under it.

ENKI is a small team that runs profitability without a finance report.

Sort the list into four groups by revenue and margin, high and low in each. That gives you four boxes: high-high, high-low, low-high and low-low. High revenue with low margin is the box that costs you most.

The fee is large enough that nobody questions the hours behind it. Name the accounts in each group before you go further.

Give each flagged account one owner, one dated action, and a target margin for the next review. Record five fields in one sheet. Client, current margin, target margin, the change agreed, and the date it happens.

An action with no named owner and no date will not survive a busy month.

Put the next review in the calendar for the same week next quarter. Re-run the three exports before it. A flagged client that has not cleared your target in two quarters means the action was wrong. Pick a different one.

You forecast client revenue and margin by month by pricing the hours you have already booked. Multiply each client’s planned hours by role against agreed billing rates. Subtract the cost rates of the people booked, then spread overhead across those months.

Carry the website build forward. Its 80 hours split across two months, 48 in July and 32 in August. The senior concept work lands first. At $150 an hour the client is billed $7,200 in July and $4,800 in August.

July costs more to deliver. Those 48 hours run at a $95 average cost rate, so delivery costs $4,560 and overhead adds $492. July nets $2,148, a 29.8% margin. August runs at $70 and nets $2,232 on smaller revenue, a 46.5% margin.

Same client, same fee, two months that look nothing alike. A quarterly average would have shown 36.5% and hidden both.

Unsigned work enters at its win probability (the chance you assign a proposal before it closes), not full value. You hold the four inputs it needs: project schedules, staffed hours by role, agreed rates, and your overhead allocation.

Productive shows forecasted profit per budget and converts your pipeline by date. A thin month appears while you can still move people onto it.

The top challenges of customer profitability analysis are getting timely data, saying no to scope creep, and forecasting accurately. Below we cover each challenge, how it manifests and how to solve it.

Getting real-time data breaks on the join. Hours live in your time tracker and invoices in your accounting tool. The two exports rarely share a client ID.

So somebody rekeys the match by hand each month, and the figures land describing a month that already closed. Fix the join before you fix the reporting. Agree one client identifier that both systems write to, and export on that key.

Then set one test for whatever setup you land on. A per-client margin figure should take minutes to produce, not a day of exports. Anything slower and nobody has the number in front of them at a renewal call.

Saying no works when the account manager has a written trigger to point at. Without one it stays a judgment call made under pressure, and the answer is usually yes.

Over-servicing, meaning unpaid work you absorb outside the agreed scope, reaches the margin long before anyone names it. Agree an hour threshold per client per month. Four hours is a workable starting point on a retainer.

Anything above it goes back as a change order naming the hours, the price, and the new delivery date. Log the requests you absorb anyway against that client. Hours nobody logs disappear into non-billable time, and the margin reads fine while the team works late.

You know the trigger works when change orders start arriving. If a quarter passes without a single one, the threshold sits too high or nobody is watching it.

The hard part of forecasting a client budget is the cost side. It keeps moving after you have committed to the price. Salaries rise mid-year, a senior stands in for a junior who left, and a subcontractor quotes high.

A fixed fee agreed in January can lose four points by June with no extra hour logged. Re-run the cost rates in your forecast whenever payroll changes.

You improve customer profitability by changing how the business prices and delivers work. Ranking your clients fixes one account at a time. These levers change the inputs that set cost and price across all of them.

With unprofitable clients you have three moves: renegotiate scope and price, cut delivery cost, or offboard the account. Flag any client sitting below your target for two quarters, then pick one move and give it an owner.

Renegotiate when the fee stopped matching the work. Say a retainer has held at the same monthly fee for two years while the deliverables doubled. Bring the logged hours and the original scope to the renewal call. Show what the fee now buys, then propose a higher fee or a smaller scope.

Cut delivery cost when the fee is fair and the team is too senior for the job. Move reporting, resizes and production work to junior staff, and keep senior hours on strategy. The account comes off your flag list once it clears target for two consecutive quarters. The same test works one level down, at project profitability per job.

Offboard when neither of the first two moves closes the gap. Give the notice your contract requires and hand over files and access in writing. Book the freed hours to the accounts at the top of your ranking.

A client ranking is only worth building if something changes on the accounts at the bottom of it. The work that pays comes next. The repricing call, the different delivery mix, the notice you finally give.

Rank your clients this quarter, pick one move per flagged account, and check the same list next quarter. If you want those numbers without three exports and a rekey, Productive keeps them in one place.

Book a demo and start today.