Owners: how much can you pay for a marketing engagement?
The question is not how much an agency costs. It is how much your margin lets you pay. The calculation fits in four lines, uses your own numbers, and gives a ceiling beyond which an engagement destroys profit whatever its results.
- Your ceiling is calculated from the sales the engagement must produce and your gross margin per sale, not from a percentage of revenue.
- The most cited benchmark in the industry, 7.8% of revenue, comes from a survey of companies with more than one billion dollars in revenue. It says nothing about your situation.
- Seven factors make the cost of an engagement vary fivefold. Five of them sit with you, not with the provider.
- A credible budget includes fees, media, tools and recurring costs. Most proposals forget the fourth item.
- An initial analysis kept separate from the monthly plan gives a readable budget: you know what you pay to understand and what you pay to execute.
On this page
What the sustainable fee ceiling is
The sustainable fee ceiling is the maximum amount a company can put toward a marketing engagement, including fees, media, tools and recurring costs, without the operation destroying profit. It is calculated from the number of sales the engagement must produce and the gross margin per sale, never from a percentage of revenue. It is a ceiling, not a target: spending less remains possible, spending more is never profitable.
Why the question is wrong
This article is for executives and management teams about to request proposals with no order of magnitude in mind. It does not give a price, it gives the way to calculate yours.
"How much does an agency cost" has no more of an answer than "how much does an employee cost." It depends on what they do, the time they spend on it and what you expect from it. The useful question is different: how much can your company pay before the operation stops being profitable?
This question has an answer, and it comes from you. A company that knows it changes its posture. It stops comparing three prices and starts evaluating three proposals against a ceiling it has calculated itself.
It also avoids the two symmetrical mistakes: paying too much for work that cannot produce enough margin, and paying too little for work that could have produced a lot.
Calculate your ceiling in four lines
| Line | What you take | Scenario |
|---|---|---|
| 1. Expected incremental sales | Sales the engagement must produce in a year, beyond what you would have obtained without it | 40 |
| 2. Gross margin per sale | Real margin, after direct production and installation costs | $8,000 |
| 3. Margin created | Line 1 multiplied by line 2 | $320,000 |
| 4. Share reinvested in acquisition | Share of that margin you agree to put back into acquisition | 45%, or $144,000 |
That $144,000 is the all-in annual ceiling: fees, media, tools and recurring costs. If you plan for $72,000 in media and $8,000 in tools and recurring costs, $64,000 remains for fees, or about $5,300 a month.
Three clarifications. The word incremental in line 1 is essential: count the sales you would not have had, not all of your sales. The method to estimate it is in our article on incrementality tests. Line 2 requires the real gross margin, returns and warranties included, not revenue. And line 4 is a management decision: 45% is a scenario, a growing company can go higher, a company under pressure will go lower.
Redo this calculation with your numbers before requesting a single proposal. It takes an hour and it changes every conversation that follows.
A final word on the most cited benchmark in the industry, which holds that a company puts around 7.8% of its revenue toward marketing. It is real, it is measured seriously, and it probably does not apply to you.
A note on scope settles the question. This figure comes from a survey of 401 marketing leaders in North America and Europe, in companies mostly reporting more than one billion dollars in annual revenue. A Quebec company of 40 employees has neither the same cost structure, nor the same channels, nor the same constraints.
The problem is not the quality of the data, it is its use. Applying a percentage drawn from an entirely different segment amounts to setting your budget on the average of a group you do not belong to.
A provider who justifies a budget by citing a percentage of revenue without stating where that percentage comes from is offering you a norm, not a calculation. Always ask which sample and which territory the figure rests on. The question alone is enough to tell a sales pitch from a line of reasoning.
The seven factors that make an engagement vary
The state of your data. A structured catalogue and recorded sales cut the work in half. A messy starting point is paid for in hours of cleanup.
The length of your sales cycle. A long cycle stretches measurement, therefore the time before you can adjust, therefore the length of the engagement.
The number of channels actually run. Each channel adds management, production and measurement.
The complexity of your product. A catalogue of 30,000 items or a configured product calls for technical understanding that cannot be improvised.
What you provide. The real questions of your buyers, your sales data, your decisions. It is the factor that most changes the real cost, and the only one you control entirely.
Inherited execution debt. A file taken over calls for catch-up that produces no gain, as explained in our article on taking over a failed engagement.
The level of measurement required. A connection all the way to the closed sale costs more than a campaign report, and is worth much more.
Five of these seven factors sit with you. That is why two companies in the same sector and of the same size receive very different proposals, without any provider being dishonest.
What a credible budget contains
Four items, and the fourth is almost always missing.
Fees, meaning the work. Purchased media, which should be separate and transparent rather than folded into a single amount. The tools, licenses and subscriptions the work requires. And infrastructure recurring costs: hosting, monitoring, technical maintenance.
The fourth item is the one that derails budgets in the second year. A tagging server, for example, is an annual recurring cost that no initial proposal mentions, as detailed in our article on server-side tagging.
Demand the breakdown of all four. A proposal that presents a single amount keeps you from knowing what you could cut on the day you have to make a trade-off.
Separating the initial analysis from the monthly plan
This is the structure we use at Falia, and it settles the budget readability problem.
The first phase is a paid analysis with recommendations. It produces the current state, the missing numbers, the plan and the priorities. It has a beginning and an end, a price known in advance, and it belongs to you even if the engagement does not follow.
The second phase is a monthly execution plan, once you know what to do and in what order. It carries the ongoing work and is reviewed on a set date.
Three advantages for you. You know exactly what you pay to understand and what you pay to execute. You can stop after the analysis without having lost anything. And you avoid the pattern described in our article on the contractual framework, where a uniform flat fee makes you pay late for work delivered early.
What a price that is too low really buys
A price clearly lower than the others is not a bargain, it is information about the content of the engagement.
It usually buys one of these three arrangements. Reduced working time, often spread across too many clients to allow a real understanding of your company. Execution without measurement, where the result will be a campaign report rather than a cost per sale. Or standardized work applied the same way to everyone, which works for a standard need and fails on a complex catalogue or a long cycle.
None of the three is illegitimate. A small company with a simple need is sometimes well served by a low-priced standardized offer. The problem appears when the need is complex and the budget matches simple work: you then pay for work that cannot produce the expected result, which is the surest way to lose everything.
What to establish before requesting proposals
These five numbers come from your sales and production data. Without them, you will compare prices instead of evaluating offers.
- How many incremental sales must an engagement produce for you to judge it a success?
- What real gross margin does your average sale carry, returns and warranties included?
- What share of the margin created do you agree to reinvest in acquisition?
- What can you provide yourself in data, buyer questions and time each month?
- If the fees exceed your calculated ceiling, do you reduce the scope or walk away, and who makes that decision?
The useful answer gives a number of sales and a margin in dollars. A weak answer talks about a budget of roughly so much a month. A provider who proposes an amount without having asked for your margin cannot know whether the engagement will be profitable for you, only whether it is profitable for them.
Establishing this ceiling with your numbers and testing it against what an engagement would really demand is precisely the object of a paid audit.
What stays with you: the real gross margin per sale, the share of margin you agree to reinvest, and the decision to reduce the scope rather than the price. What gets delegated: estimating the incremental sales, breaking a budget down into four items, leveling up the proposals received and calculating the ceiling. A company that arrives with its ceiling calculated gets proposals adjusted to its reality and decides within a week. A company that asks for prices with no ceiling receives three amounts with no benchmark, picks the middle one, and discovers after a year that the scope did not match the problem.
The framework for reading a proposal is in how to read an agency proposal.
Building the budget that carries these decisions is detailed in splitting the budget across channels.
Structuring growth work over several years, with a budget that holds up, is at the heart of the Generate demand and growth goal.
Want to see how we structure our engagements before discussing one? See our way of executing.
Frequently asked questions about the cost of an engagement
Why not simply use a percentage of revenue?
Because a percentage ignores your margin, which is the only thing that determines what a sale actually earns you. Two companies with the same revenue, one at 15% margin and the other at 40%, cannot sustain the same budget. The most cited benchmark also comes from companies with more than one billion in revenue, which makes it hard to transpose to an SMB.
How do you estimate incremental sales before starting?
Through an assumed order of magnitude rather than a forecast. Look at what your current sources produce, and ask yourself what serious work would realistically add over twelve months. The estimate will be imperfect, and it is always better than no benchmark. It is then corrected with real measurement.
Should the ceiling include the media budget?
Yes, along with the tools and recurring costs. The ceiling covers the total cost of acquisition, not fees alone. That is what then lets you make a trade-off between paying for more work and buying more media, a trade-off that only makes sense if both come out of the same envelope.
What if all the proposals exceed my ceiling?
Two options, and only one is good. Reduce the scope to bring it under the ceiling, accepting to do fewer things better. Or walk away for now and first work on what costs nothing in media, such as the callback time or qualification. Negotiating the price at constant scope gives underfunded work, which is the worst of the three outcomes.
Why pay for an analysis before the engagement?
Because it lets you decide with full knowledge, including not to proceed. A provider who gives their analysis for free necessarily funds it within the engagement, and has an interest in their conclusions leading to a sale. A paid analysis belongs to you and can conclude that nothing should be done this year.
Does the ceiling change over time?
Yes, with your margin and your capacity. A company nearing its production ceiling no longer needs the same acquisition volume, which mechanically lowers its useful fee ceiling. Recalculate it once a year, at the same time as the review of the contractual framework.
- Gartner, Gartner 2026 CMO Spend Survey, press release of 11 May 2026. Survey of 401 marketing leaders in North America and Europe, in companies mostly reporting more than one billion dollars in annual revenue. Source of the 7.8% of revenue figure.
- Falia working framework, calculation of the sustainable fee ceiling and typology of the seven factors of variation. The amounts are explicit scenarios, to be redone with your margin and your volumes.
Gabriel almost always takes your first call and carries out your audit. He builds the strategy starting from your growth goal: where to put your budget, which market to test and how to connect each lead to a real sale in your CRM. He mainly leads engagements for three goals: Optimize the profitability of your digital campaigns, Develop a new market, and Generate demand and growth. With Geneviève, he also works on organic search (SEO), AI visibility (GEO) and conversion rate optimization (CRO). The sales a Google Ads or Meta Ads campaign brings in depend on the page that receives the click. He writes mainly about marketing strategy, paid advertising and measurement.
About Falia →
SEO
Paid advertising
Conversion
AI visibility
Web design