Retainer vs Project-Based Pricing for Agencies

Retainer vs project pricing: what is the difference?

A retainer is a recurring fee for ongoing access to a team or a defined volume of work each month. Project pricing is a one-time fee for a specific deliverable with a defined scope, timeline, and acceptance criteria. Retainers trade unit margin for revenue predictability and deeper client relationships; project pricing trades predictability for cleaner boundaries and higher margin per engagement.

The distinction matters because the two models put risk in different places. In a retainer, the client carries volume risk: if they need more than the included hours, they pay overage or wait. In a fixed-fee project, the agency carries delivery risk: if the work takes longer than estimated, the agency absorbs the overrun.

Industry surveys consistently show most agencies run both. Promethean Research’s 2026 digital services research found roughly 95% of agencies offer project work and 91% offer retainers, with 88% offering both. The model choice is rarely either/or; the interesting question is the mix and the pricing discipline inside each.

How do you price a retainer so it stays profitable?

Price a retainer from delivery cost, not from a round number that feels acceptable in a sales call. The reliable formula is retainer price = team cost ÷ (1 − target margin), where team cost is fully loaded (salary plus about 30% for benefits, tools, and overhead) and divided by realistic billable hours, not 2,080 per person.

Three inputs decide the result, and agencies usually get all three wrong in the same direction.

1. Hours actually consumed, not hours you wish were consumed. Include account management, reporting, and review calls. A 40-hour design retainer typically carries 6–10 hours of non-production work.
2. Fully loaded cost per hour. Take annual salary × 1.30, divide by 2,080, then adjust upward because nobody is billable 100% of the time. Parakeeto’s benchmark work puts healthy agency-wide net utilization at 50–60%, with delivery staff at 65–85%. If you priced at 100% utilization, you priced a fantasy.
3. Target margin. A 50–65% gross margin on delivery is the usual range for a services business that also carries sales, admin, and bench time.

Worked example: building a $15,300 monthly retainer

A small agency staffs a retainer with a strategist at $95,000 and a developer at $120,000. Using the 1.30 loading factor:

• Strategist: $95,000 × 1.30 = $123,500 ÷ 2,080 = $59.38 per hour
• Developer: $120,000 × 1.30 = $156,000 ÷ 2,080 = $75.00 per hour

The scope calls for 40 strategist hours and 60 developer hours per month:

At 100 included hours, the effective rate is $153 per hour. Now add the creep that almost always appears: 20% more hours means 120 hours, and the effective rate falls to $127.50 — a 17% pay cut with no change in the invoice.

That is why the contract needs a cap and a rate: “up to 100 hours per month; additional hours billed at $165 per hour, approved in writing.” Without that sentence, the retainer is an unlimited subscription to your team.

Should unused retainer hours roll over?

Roll over a limited amount, and only for a limited time. Unlimited rollover turns your retainer into a liability that grows every quiet month, then lands in one impossible sprint. A workable policy is 25% rollover expiring after 60 days. If a client never uses their hours, the retainer is oversized — right-size it at renewal.

How do you price a fixed-fee project without gambling?

Price a fixed fee by estimating hours bottom-up, then adding an explicit risk buffer that reflects how much you actually know. The common sequence is: task-level estimate → three-point adjustment for uncertainty → contingency of 10–25% → target margin → price. Skipping the contingency is not optimism; it is transferring your estimation error to your own P&L.

Estimation quality improves fast when you keep a quote-versus-actual log. Track estimated hours, actual hours, and the reason for variance on every project. After ten projects you will know that your “discovery” phase runs 40% over on enterprise clients and your “QA” phase runs 15% under on marketing sites.

Worked example: a $62,000 website build

A ten-week site build is estimated at 320 hours with a blended cost of $110 per hour.

Now stress-test it. If the project runs 400 hours instead of 320, cost becomes $44,000 and margin drops to 29%. If it runs 464 hours — a 45% overrun, which happens on projects with unclear content responsibilities — cost hits $51,040 and margin collapses to 17.7%. The fee did not change. Only your profit did.

Two contract mechanics protect that margin: a written scope with explicit exclusions, and a change-order process with a rate card agreed before work starts. PMI’s guidance on fixed-price contracts is blunt about this — use change orders for scope changes rather than raiding contingency, because contingency exists for identified risk, not for client requests.

Which model produces better margins?

Project pricing usually produces higher gross margin per engagement, while retainers produce steadier margin over time and a more valuable business. Project work can reach 50–60% gross margin when estimates are accurate; retainers typically land in the 40–55% range because the price is anchored to an ongoing relationship rather than a deliverable.

But margin percentage is not the only measure. Compare the two on the numbers that reach your bank account:

The comparison is deliberately unfair — you would run four or five projects a year, not two — and that is the point. Project revenue only competes when the pipeline is full. A single slow quarter exposes the difference between a business with a recurring base and a business that restarts at zero every month.

There is also a valuation effect. Buyers pay a premium for recurring revenue because future income is more certain. An agency with 70% retainer revenue typically commands a higher multiple than one with 70% project revenue, which matters if you ever intend to sell, raise, or borrow against the business.

What does a hybrid model look like in practice?

A hybrid model pairs a retainer for recurring work with project pricing for defined builds, and it is the structure most agencies eventually adopt. The practical target is 60–70% of revenue from retainers and 30–40% from project work: enough recurring income to cover payroll, enough project upside to grow.

Here is what that looks like at $100,000 of monthly revenue.

The design rule is simple: retainers pay for capacity, projects pay for profit. If your retainer base covers 100% of payroll, you can decline bad-fit projects. If it covers 40%, every slow month becomes a cash crisis and you start discounting to fill the gap — which is how agencies end up at 12% net margin wondering where the money went.

A second rule: never let a project hide inside a retainer. When a client asks for “a quick landing page” as part of their monthly support, either it fits the included hours or it becomes a separate statement of work. The moment you absorb it, you have taught the client that your scope document is decorative.

What are the trade-offs of each model?

Both models have a failure condition, and the honest comparison names it.

Retainers give you forecastable cash, lower sales cost per dollar earned, deeper client knowledge, and a more sellable business. They also cap your upside, invite scope creep, create unused-hours guilt on both sides, and make you dependent on a small number of relationships. Losing one anchor retainer can remove a third of revenue overnight.

Project pricing gives you clean boundaries, a defined finish line, the ability to reprice with each engagement, and often better margin. It also gives you feast-famine revenue, repeated sales cycles, estimation risk that lands entirely on you, and a temptation to say yes to badly scoped work because the pipeline looks thin.

Hybrid gives you the stability of one and the upside of the other at the cost of administrative complexity. You need two sets of contracts, two reporting rhythms, and the discipline to keep them separate. Agencies that blur the line end up with retainer pricing on project scope — the worst of both.

What mistakes erode margin in both models?

• Pricing from a round number instead of cost. $5,000 a month feels normal because other agencies quote it. It says nothing about whether you can deliver it profitably.
• Assuming 100% utilization. Bench time, sales support, and internal work are real costs. Price them in or plan to absorb them.
• No written scope. Verbal scope is an open invitation to creep. Document deliverables, revision rounds, and exclusions in writing.
• No change-order process. If the only way to say no is a difficult conversation, you will say yes. A rate card makes it routine.
• Tracking revenue but not per-client profitability. A client can be 20% of revenue and 2% of profit. You will not know without hours, costs, and fees in one view.
• Never raising prices. Costs rise every year. Agencies that avoid annual reviews quietly give away 3–5% of margin annually.
• Discounting to win a retainer you cannot staff. A retainer sold below delivery cost is not a foot in the door; it is a subsidized job.

> Common mistake: Treating the retainer fee as revenue and the team’s time as free. The fee is only meaningful next to the delivery cost of the hours actually consumed.

Which clients should get which model?

Match the model to the shape of the client’s demand, not to how much they can pay.

The last row is the one agencies ignore most often. If a client cannot describe what “done” means, a fixed fee is a bet against your own estimate. Sell a two-week paid discovery, then price the build with real information.

How do you track per-client profitability without a finance team?

You need three numbers per engagement: hours consumed, fully loaded cost of those hours, and fees invoiced. That is it. A spreadsheet handles it for a client base under fifteen, provided someone updates it weekly and the hours come from the same place the work is planned.

The usual failure is fragmentation. Time lives in one tool, the budget in another, and the invoice in accounting. By the time the monthly review happens, reconstructing the picture consumes an afternoon and produces numbers nobody trusts — so the review gets skipped and the creep becomes permanent.

Doitify is built for the version of this problem where the work and the numbers should live together: tasks and sub-tasks with owners and estimates, workload views per person, and work and performance reports that show planned versus actual on the same board. It is not an agency ERP and does not replace your accounting system; the value is that the delivery signal and the budget signal come from the same record. We build it, so treat that as a product note rather than an independent review.

Conclusion

Retainer vs project pricing is not a personality test for your agency. It is a decision about where you want risk to sit and how much predictability you need to sleep at night.

Price retainers from loaded cost and realistic utilization, cap the hours, and revisit the number every year. Price projects with a real risk buffer and a change-order process you are willing to use. Then track effective hourly rate per client, because that single number exposes both a retainer that has quietly become a discount and a project that looked profitable at signature.

Get those mechanics right and the model argument mostly disappears. Agencies rarely struggle because they chose the wrong model; they struggle because they priced once and never looked again.

If this post on retainer vs project pricing was helpful, you might also enjoy How to Calculate the Critical Path (Step-by-Step) and What Is Resource Management in Project Management?.