You asked for two proposals and got back two documents you cannot compare.
One charges a flat monthly amount. The other charges a percentage of your ad spend. One says "media management". The other says "growth". Neither one says what a new landing page costs, and both want a meeting before they talk numbers.
I was on that side for years. I own a kids entertainment business inside shopping malls, I am not an agency and not a consultant. I paid different vendors to handle my website, my traffic and my social media, and it took me far too long to understand that the right question was never "is this expensive?".
This piece is about how to read an agency proposal with the eyes of the person paying the bill.
Why nobody publishes a price list
You have probably noticed that almost no agency puts prices on its website.
It is not a conspiracy. There are three real reasons.
The first is that the work genuinely varies. A local business with one offer and a business with thirty products are different amounts of work.
The second is commercial: published prices allow comparison, and comparison pushes value down. With no public list, each proposal is negotiated in isolation, and the price adjusts to the apparent size of your business.
The third one matters most to you. Price only means something next to scope, and scope in marketing is elastic. "Social media management" can mean twelve posts a month or it can mean two. Both fit on the same line of the proposal.
That is why "how much does an agency cost" has no single answer. The question that does have an answer is a different one: what exactly is being delivered, and what is left in my hands at the end?
The pricing models, and the incentive hidden in each
Practically every proposal falls into one of three shapes, or a mix of them.
Flat monthly retainer. You pay a set amount per month for a package of services. It is the most common and the easiest to budget for. The thing to watch is that the amount does not move when results move, in either direction. The agency's incentive is to keep you satisfied, which is not the same thing as making you sell more.
Percentage of ad spend. You pay a percentage of what you invest in media. It sounds fair, and it is common on larger accounts. The uncomfortable detail: the vendor earns more when you spend more, not necessarily when you profit more. That does not imply bad faith, it means the interests are not identical.
Fixed project. A website, a campaign, a batch of creatives, with a defined price and deadline. It is the most transparent of the three. It is also the one that tends to leave you alone the day after delivery, when the project ends and the operation keeps going.
Plenty of proposals combine all three: a monthly fee, plus a percentage above a certain spend, plus projects billed separately. Read it twice to find out which part grows without you noticing.
What is usually inside the fee, and what becomes an extra
This is the line where the bill fattens after you sign.
Usually inside: campaign management, reporting, a few creatives per month, small tweaks to the site, periodic meetings.
Usually an extra: video production, photo shoots, a new landing page, a campaign page, structural changes to the site, third party paid tools, ad spend, boosting fees, and anything that shows up "out of scope".
None of those extras is abusive on its own. The problem is finding out afterwards. Before you sign, take the proposal and ask item by item: is this included, or is it billed separately? Ask for the answer in writing, in the same document.
Ask that question especially about landing pages. It is the piece you will most want to create and replace, because every new campaign wants its own. If each page costs an extra, your marketing will run slower than it needs to, for price list reasons.
Ad spend is not the cost of the agency
This confusion is so common that it deserves its own paragraph.
There are two different pools of money in your account. One goes to the agency, for its work. The other goes to Meta and Google, to buy impressions and clicks.
The second one is yours either way. It still exists if you switch agencies, and it still exists if you take the operation over yourself. It is not a possible saving, it is media investment.
Mixing the two numbers leads to two bad, opposite decisions. One is thinking marketing costs a fortune when most of it is budget that turned into customers. The other is thinking the agency is cheap because its fee is small next to the spend, without noticing that the fee is fixed and the spend is variable.
Split the two lines in your spreadsheet, now. Everything else in this analysis depends on it.
The costs that never appear in the proposal
There is a second bill, written in no document at all, and it is usually bigger than we like to admit.
Waiting. How many days between asking for a change and seeing it live? Multiply by how many changes you request per month. That is opportunity sitting still.
Rework. The piece comes back off tone, you explain again, it comes back again. Every round is your time, and an owner's time is the most expensive resource in the business.
Approvals. Meetings, emails back and forth, group chats with three different vendors who never talk to each other. You become the cable connecting everyone.
Translation. You know the customer. Whoever executes does not. The distance between those two things gets filled with briefings, and a briefing is you working for free.
Add up the hours. Not to get angry, but to have the number in hand when someone tells you that taking over your own marketing "is a lot of work". The honest comparison is not between working and not working. It is between two kinds of work.
The question that reveals the real price: what did each new customer cost?
Everything else is a detail next to this.
Take any month. Add what you paid in fees and what you paid in ad spend. Divide by the number of new customers who came from those channels in that period.
The result is your cost per customer. It is the only number that answers whether the arrangement is worth it.
Two things will happen when you try to do this math.
The first is that you probably will not manage it on the first try, because you do not know how many new customers came from ads. That is not your failure. That is missing measurement, and it is the clearest symptom that the data is not organized in your favor.
The second is that, once the math is done, the conversation changes subject. It moves from "this is expensive" to "each customer costs this much and leaves me this much". Now you have a business decision instead of a feeling.
If you cannot reach that number today, that is the problem to solve before you replace any vendor.
What stays with you when the contract ends
Here is the point that changed my mind, and it has nothing to do with price.
Run the thought experiment: you cancel today. What still exists tomorrow?
Does the site stay online, on your domain, or was it hosted in the agency's account? Is the Meta Pixel, the code that records who visited your site and lets you advertise to those people again, in an account of yours or in one of theirs? Who keeps the campaign history, which is what makes the next campaigns perform better? Is the contact list you captured sitting in a tool registered in your name?
If the answers are "theirs", you were not buying marketing. You were renting it.
And then the price means something else. Paying a lot for something that stays is an investment. Paying a lot for something that evaporates when the contract ends is rent, and rent you pay forever.
The problem was never the size of the monthly bill. It was finishing five years of payments with no asset in my own name.
How to read your own proposal with these eyes
Take the document on your desk and run it through this list.
- What is the pricing model? Flat, percentage, project, or a mix. If it is a mix, which part grows on its own?
- What is included, item by item? Ask for the closed list in writing, especially about new pages.
- Is ad spend separated from the fee? If they are in the same number, ask for them to be split.
- Whose name is on the domain, the hosting, the pixel, the tag and the ad accounts? Ask for the answer in writing. It is the most important question on this list.
- What is the average turnaround for a change request? And what happens when it slips.
- How do I access the data without asking? If the answer is "we send a report", you do not have access, you have a courtesy.
- What happens at termination? What the notice period is, and what gets transferred to you.
None of these questions is aggressive. A good vendor answers all of them without hesitating, and some answer better than you expect. The bad answers are valuable information too, and they cost less today than they will a year from now.
The other side of the math
This piece covered one side only: what you pay today and what you get for it.
The comparison is only complete with the other side, which is the real cost of running marketing yourself. It is not zero, and anyone who says it is has something to sell. There is a domain, an AI subscription, ad spend, and your time.
Both sides are laid out here:
- What it costs per month to run a site, landing pages and tracking without an agency, the other model's math, line by line.
- Agency or AI: the honest comparison for whoever is deciding this month, the two side by side, including the cases where an agency is still the right call.
And if you want out of theory before deciding anything, both sides of that math sit on top of one number you probably do not have written down: what you pay today to win each new customer. The free Cost Per Customer X-Ray, delivered by email, works it out with you. You paste the prompt into any free AI, answer a few questions about your spend and your new customers, and in about five minutes you have the figure. It is the number no agency ever hands you, and it turns this decision from a feeling into arithmetic.