The complaint I hear most from company owners: "we paid an agency for a full year and saw no result." And when we get into the detail, we find the problem did not start after the contract, it started in the first meeting, when the agreement was made on something that was never clear.
Agencies usually do not fail because they are not good. They fail because nobody asked the right questions before paying, so nobody knew what success even meant in that arrangement.
What follows is exactly what I tell any owner who consults me before signing a marketing contract.
1. Who will actually work on my account?
The meeting you attend has the agency owner or the account director in it, their best people. After you sign, the work goes to a different team you have never met, who may still be learning.
Ask by name: who writes the content? Who runs the ads? How many other accounts do they carry? And who is my point of contact when something goes wrong?
If the answer is vague, this is not an administrative detail, it is the single biggest factor in the result you get. The difference between someone carrying ten accounts and someone carrying three shows up in every detail, from response speed to the quality of the thinking.
2. Show me work for someone like me
A portfolio full of big brands is not evidence they will work for you. A big brand gets results from its name and its budget, not necessarily from the agency's work. And in many cases the agency executed a small part of that work and presents it as if it were all theirs.
Ask for a case from a company your size, with a budget close to yours, ideally in the same market. And ask about the numbers before and after, not the number of likes.
One question reveals a lot: "can I speak to one of those clients?" A confident agency has clients willing to talk. One that gets flustered by that question is telling you something important without meaning to.
3. What exactly are we paying for?
The difference between one proposal and another is not the price, it is what the price includes. Agree the deliverables in numbers: how many designs, how many videos, how many campaigns, ad management or execution only, who pays the ad budget, and who is responsible for replying to messages.
Also define what is not included: photography, paid media, translation, replies outside working hours. Those are the lines that turn into extra invoices two months in.
Any item not written into the contract will become a dispute within two months. No exceptions.
4. How will we measure success?
If the answer is "we'll raise brand awareness," that is not a goal, that is talk. A goal is: how many prospects a month, at what cost per prospect, and how we calculate it.
And before you agree that number, you need to know how much you can afford to pay to acquire one customer. That figure comes from your margin and the lifetime value of a customer, without it, any target set is guesswork.
Agree a realistic evaluation period up front. Ads need weeks to stabilise, and content needs months. But "we need time" with no interim numbers is not an answer.
5. Who owns the accounts and the data?
This is the point that hurts most at separation. The ad account must be in your company's name, your pages are yours, the customer data collected is yours, and reports must live somewhere you can reach.
I have seen many companies lose their entire advertising history because the account sat on the agency's email. That history is not a detail, it is what the platforms learn from, and losing it means starting from zero at a higher cost.
Write that clause into the contract explicitly: all digital assets belong to the client and are handed over within a week of the contract ending.
6. How will we communicate?
Most disputes start here rather than with the results. Agree on a written monthly report, a call every two weeks or every month, and a single channel for day-to-day contact.
And define the expected response time. An agency that takes two days to answer a simple question will have you chasing them under pressure, and that is a hidden cost on your own time.
Signals that should stop you
- A guaranteed result. Anyone promising you specific sales before knowing what your business sells for and where it loses money is selling you words.
- A price far below the market. A cheap price is covered from one place: the number of hours spent on your account.
- A one-year contract with no exit. Ask for a trial period or a termination clause with notice. A confident agency will not fear that.
- They ask you no questions. Anyone who does not ask about your margin, your ideal customer and what you have tried before will not be able to market for you.
- A proposal ready the same day. A proposal written before they understood your business is a template with your name dropped into it.
Big agency, small agency, or freelancer?
All three work, but for different situations. A large agency gives you stability, a full team and systems, but your small account will sit at the bottom of the priority list and be delivered by their least experienced people.
A small agency gives you more attention and flexibility, and its risk is dependence on one or two people, if they leave, quality drops.
A freelancer is cheaper and faster in a single specialism, and weaker on coverage: they will not do strategy, design, ads and reporting at the same level.
A practical rule: if your monthly budget is less than the salary of a mid-level marketer, a specialist freelancer or a small agency fits best. If it is much higher and you have multiple products and branches, the larger entity earns its cost.
Your own responsibility in the arrangement
Half the problems I see are caused by the client, not the agency: incomplete information, approvals that take two weeks, a change of direction every month, and replies to customers so slow that the campaign is wasted.
Before you hold them to account, make sure you have delivered your side: product and pricing information, a named person responsible for approvals, and fast responses to prospects.
An agency can bring the customer to the door of the shop. What happens after that is on you.
The first three months: what to expect
The first month is usually preparation: understanding the product, reviewing the accounts, building the first materials, and small tests. Anyone demanding sales in month one pushes the team to chase quick wins that will not last.
The second month brings the first real numbers: cost of reach, cost per prospect, and which message worked. That is the month direction gets decided.
The third month should show a clear improvement on month two's numbers. If there is no improvement and no convincing explanation, that is the time for a serious review, not for more patience.
How to compare three proposals
Build a simple table with four columns: deliverables, monthly price, method of measurement, and termination terms. You will find that proposals which looked similar are very different in the details.
And do not compare on price alone. Compare on expected cost per result: a proposal at fifteen hundred that brings ten customers is cheaper than one at seven hundred that brings two.
In short
Choosing an agency is a purchase decision like any other in business: ask, request evidence, write the agreement, and define how you will know whether you are winning or losing.
The meeting where you ask hard questions is the cheapest meeting you will hold this year. And a good agency will not resent the questions, they will be relieved to be working with someone who knows what they want.
This is part of a series on choosing a marketing agency, which also covers why an agency cannot write your business plan, and how to build a marketing team.