How a Marketing Agency Found Out Which Retainers Were Losing Money

How a Marketing Agency Found Out Which Retainers Were Losing Money

This is an illustrative, anonymized composite drawn from patterns common to agencies at this size. It is not a specific client engagement, figures are rounded, and outcomes vary. Nothing here is a promise of a result.

Agency client profitability is the number most agency owners think they have and almost none actually do. They have a company-wide gross margin, which is an average, and an average made of winners and losers tells you nothing about either.

The agency in this composite does $4 million in revenue, has 22 people, and had watched gross margin slide from 54 percent to 50 percent over two years while revenue grew. Every explanation offered in the leadership meeting was plausible. None of them were checkable.

Why One Blended Margin Hides Agency Client Profitability

A single margin number describes a company that does not exist. It describes the imaginary average client.

Real agencies have a spread. Some retainers run at 65 percent gross margin because the scope is tight and the team is efficient. Some run at 20 percent because scope grew for three years and price did not. Blend them and you get 50 percent, which is a number that leads to exactly the wrong decision: raise prices across the board, or cut costs across the board, when the actual problem is concentrated in three accounts.

The owner's first instinct was a hiring freeze. That would have punished the profitable accounts to subsidize the unprofitable ones.

The Change That Made It Visible

Almost every agency at this size has the same accounting structure problem: all payroll sits in one operating expense bucket.

That single choice makes per-client profitability impossible. If the people delivering the work are in overhead, there is no cost of services to subtract from client revenue, so there is no gross margin by client to calculate. The fix is to split payroll between delivery and everything else, putting client-facing delivery labor into cost of services alongside contractors, freelancers, and pass-through media.

It sounds like a bookkeeping detail. It is the whole thing. Once delivery labor is in cost of services, the P&L can be cut by client, and the average stops hiding the outliers.

Two honest caveats. Splitting payroll requires a defensible allocation basis, usually hours, and reclassifying prior periods means your year-over-year comparisons need restating before anyone reads them. Neither is hard. Both take a cycle.

The Time Tracking Nobody Wants to Do

The allocation needs hours, and agencies hate tracking hours. The objection is real: creative people resent it, the data is often fiction, and precision beyond a point costs more than it returns.

The version that works is deliberately loose. Track to the client, not the task. Round to the half hour. Do it weekly rather than daily. Accept 85 percent accuracy, because the decision you are trying to make is whether an account is at 20 percent margin or 60 percent margin, and no reasonable amount of tracking error changes that answer.

Where this fails is agencies that use the data punitively. Once hours become a performance metric for individuals rather than a costing input for accounts, the numbers stop being usable within a month.

What the Numbers Showed

Running six months of restated data produced the pattern that shows up in most agencies that do this exercise for the first time.

The spread was much wider than anyone guessed. Several accounts landed well above the blended average, most clustered near it, and three sat far below the target margin. Two of the three were small accounts that had quietly accumulated scope. The third was the largest logo on the website, an account everyone assumed was carrying the agency, which was running near break-even after delivery labor.

That last one is the common result, and it is the uncomfortable one, because the marquee client is usually the account with the most scope creep and the least willingness to be repriced.

Reprice, Rescope, or Release

Once the data exists, there are only three levers, and each carries a cost.

Reprice. Go back with a new rate tied to the actual scope. It is the cleanest fix and the one with real churn risk, because some clients will leave and the ones most likely to leave are the ones already paying below market.

Rescope. Keep the price and cut the work back to what was originally sold. Lower conflict than repricing, but it requires the discipline to say no to out-of-scope requests, which is the discipline that created the problem.

Release. Give the account back. This looks like losing revenue and often is, in the near term. It only makes sense when the account is genuinely below variable cost or is consuming senior capacity that has a better use, and it depends on being able to redeploy or reduce the team, which is not instant.

Most agencies use all three across different accounts. What they cannot do is apply one of them uniformly, which is exactly what a blended margin pushes them toward.

The Realistic Range of Outcomes

In composites like this one, agencies that reprice or rescope their bottom accounts typically recover a portion of the margin they lost, over two to four quarters rather than in a quarter.

The range is wide and depends on client concentration, contract terms, notice periods, and how much of the delivery cost is fixed salary versus flexible contractor spend. Some agencies recover most of the gap. Some lose an account and land roughly flat on revenue with better margin and less stress. Some find the low-margin accounts are strategic for reasons the P&L cannot see, such as reference value or category credibility, and choose to keep them with eyes open. That last outcome is a legitimate result of the exercise, not a failure of it.

What reliably changes is the quality of the decision. An owner arguing about a hiring freeze without per-client margin is guessing. The same owner with the data is making a choice.

Who Does Client Profitability Work for an Agency

This sits above bookkeeping and below a full-time CFO, which is the gap fractional CFO engagements exist to fill. It needs someone who can restructure the chart of accounts, build the allocation, and then sit in the room when the recommendation is to reprice the biggest client.

8 Figure Finance is a Philadelphia firm founded in 2024 providing CFO, accounting, and tax work specifically for marketing agencies doing $1 million to $20 million, which is the exact band where this problem appears and the structure to fix it usually does not exist yet.

8 Figure Finance has 33 verified client reviews on Sam's List as of 2026-08-06. Reviews reflect the experiences of individual clients, do not represent an endorsement by Sam's List, and are not indicative of future results.

No advisor can make a client accept a price increase, and restating prior periods can surface numbers you have already reported to a lender or a partner, which is a conversation worth planning for rather than discovering. Ask any firm what the first 90 days produce and what it will require from your team, since this work fails when the agency cannot supply the hours data.

You can compare firms by specialty and verified reviews in the Sam's List fractional CFO directory.

Frequently Asked Questions

How do you calculate client profitability at an agency? Take the revenue from each client and subtract the direct cost of delivering it, which means delivery payroll allocated by hours, contractors, freelancers, and any pass-through costs. The result is gross margin by client. This is only possible if delivery labor sits in cost of services rather than in general overhead.

What is a good gross margin for a marketing agency? Agency benchmarks commonly cited fall in the 50 to 60 percent range once delivery labor is treated as a cost of services, but the useful number is your own spread across accounts rather than a benchmark. An agency at a healthy average can still have accounts losing money inside it.

Should an agency fire an unprofitable client? Not automatically. Repricing or rescoping usually costs less and preserves the relationship. Releasing an account makes sense when it is below variable cost or consuming senior capacity with a better use, and only when the team can be redeployed or reduced, which takes time.

How much time tracking does per-client profitability require? Less than most agencies fear. Tracking to the client rather than the task, rounded to the half hour and logged weekly, is generally accurate enough to distinguish a 20 percent margin account from a 60 percent one. Precision beyond that rarely changes the decision it informs.

---

About the author: Kimberly Green is the cofounder of Sam's List, where business owners and high earners find vetted CPAs, financial advisors, and fractional CFOs. She's met one-on-one with 400+ financial professionals and writes from the real data behind thousands of client-advisor matches. Ask her anything about finding an accountant - she's heard it all, including the questions people are afraid to ask.