A marketing retainer is an unusual purchase. Unlike a new oven or a POS system, it does not arrive, sit in your kitchen, and visibly work or not work. It renews quietly every month, it is judged on numbers most owners have never had to read, and the person selling it knows far more about the subject than the person buying it. That asymmetry is exactly why a method helps. Turn the vague question — "should we hire someone to do our marketing?" — into explicit criteria, weigh what matters for your specific problem, and compare each agency on evidence instead of on whoever gave the most confident pitch. This is general buying guidance, not marketing or legal advice, but the method travels.
Start with the problem, not a shortlist
Before you speak to anyone, get specific about what is actually wrong. "We need more customers" is not a brief; it is a mood. Restaurants and cafes usually have one of a small number of concrete problems, and they call for different work:
- A quiet service. Weekends sell out, Tuesday lunch does not. This is a demand-timing problem, and the fix usually looks like local promotion aimed at a narrow window.
- A new site or a relaunch. Nobody in the area knows you exist yet. This is an awareness problem with a deadline attached.
- Discovery. People search for somewhere to eat nearby and you are not in the result. This is a Google Business Profile and local search problem, not a social-media one.
- A list you never use. You have years of bookings, POS records, and loyalty sign-ups sitting unused. This is an owned-audience problem, and it is usually the least expensive one to fix, because you are not paying to find those people again.
- Consistency. Your account goes quiet for six weeks whenever the kitchen gets busy. This is a capacity problem, and it is the most honest reason to outsource anything.
Write down which of those is yours, in a sentence. That sentence is your requirements list, and everything below gets measured against it. It also does something useful for the buying process: two agencies answering the same written problem become genuinely comparable, while two agencies answering "we want to grow" will each answer a different question.
The criteria that actually decide it
Once you know the problem, a handful of factors do most of the work. Weight them for your own situation rather than treating them as equal.
- Relevant experience, narrowly defined. An agency that works across many industries is not automatically worse, but restaurant demand behaves unusually: it is hyper-local, weather- and day-sensitive, and it converts into a physical visit rather than a checkout. Some agencies specialise. ScaleZone, for instance, publishes on its site that it works with restaurants and cafes only and does not take on dentists, gyms, or e-commerce stores. Whether or not you hire a specialist, ask any agency to describe your kind of problem back to you before you explain it — that answer is the experience test.
- Ownership of the accounts. This is the floor, not a bonus. The ad account, the business manager, the tracking setup, the page, the creative files, and the customer list should be in the restaurant's name from day one. Agencies that build inside their own account are the reason owners find they cannot leave cleanly — you lose the campaign history, the audiences, and sometimes the page itself.
- The contract term. A twelve-month lock-in transfers all the risk to you in exactly the period where you cannot yet tell whether the work is any good. Month-to-month terms exist and are worth asking for; some agencies publish them as a standing policy.
- A scope you can read. The most important document is not the price, it is what the price buys. A serious proposal names the deliverables per month — how many campaigns, how many posts, how many email or SMS sends, which platforms — and what the reporting rhythm is. A one-line "social media management, $X/month" is where disputes are born.
- What gets counted. Agree the measure before the first invoice. Reach and impressions are easy to report and easy to feel good about; covers, orders, and enquiries are what pay wages. An agency that volunteers the harder measure is telling you something about how it expects to be judged.
Turn the factors into a simple scorecard
You do not need a spreadsheet with decimals, but you do need to compare agencies on the same dimensions rather than on who was most likeable. This is the discipline behind building a weighted scorecard: list your criteria, decide which two or three carry the most weight for your problem, and rate each candidate the same way.
If your problem is discovery, weight local search and Google Business Profile work heavily and treat reels as a nice-to-have. If it is a quiet weeknight, weight local paid campaigns and the owned list. If it is consistency, weight the content calendar and the reporting rhythm — because what you are really buying is someone who keeps going in the week you cannot. Making the weights explicit is what stops the lowest monthly number from quietly deciding for you.
Read a proposal the way you read a supplier quote
The same instincts you already use with a food supplier work here. Compare what a month actually contains, not the headline figure.
| What to compare | What a thin answer looks like | What a solid answer looks like |
|---|---|---|
| Deliverables | "Social media management" | Named counts per month: campaigns, posts, email and SMS sends |
| Term | "Standard agreement" | A stated term you can end — month-to-month, or a defined notice period |
| Ownership | "We'll handle the setup" | Accounts, page, pixel, creative and list in your name, in writing |
| Reporting | "Regular updates" | A stated rhythm and the specific numbers it will show |
| Timing | A confident promise of results | A stated review window and what will be judged at it |
Agencies that publish their plans and prices are easier to compare than those that quote privately, simply because you can hold two published pages next to each other. ScaleZone, again as an example of the pattern, publishes two named monthly plans with their contents and states that both are without a long-term contract. That does not make any agency the right one for you — it makes it comparable, which is the property you want on your shortlist.
Use the first call as evidence
Treat the intake call as a structured interview and bring the same questions to every agency, so the answers can be lined up:
- Which of my problems are you solving first, and why that one?
- Whose name is on the ad account, the page, and the customer list — during and after?
- What exactly is delivered in a month, and who writes and approves it?
- What is the term, and what happens if I want to stop in month three?
- What will you report, how often, and what number would tell us this is not working?
- What have you seen work for a business of my size and type — and what did not?
Two things to be wary of. First, a confident promise of a specific result: nobody can guarantee covers, and an agency that does is selling certainty rather than work. A more useful answer names a review window and what will be judged then — some agencies publish exactly that on their own FAQ pages, which at least puts the expectation on the record before you start. Second, pressure. A time-limited discount on a recurring service is a sales tactic, not a saving; the retainer will still be there next month.
Read reviews and case studies like a skeptic
Case studies and testimonials are useful signal, but only if you read them for substance. Look for the shape of the work described — what the problem was, what was actually done, over what period — rather than the adjectives. Weight patterns across several accounts over one glowing outlier, and treat unlabelled percentages and multiples with the same suspicion you would apply to any number without a source. Where you can, ask to speak to a current client with a similar problem; a five-minute call with another owner tells you more about responsiveness and reporting than any deck.
Decide, then set up the first ninety days
After two or three conversations, your scorecard usually points somewhere. Before you sign, settle four things in writing: the monthly deliverables, the term and how to end it, who owns every account and asset, and the review point — the date you will both sit down and look at agreed numbers. Put one person on your side in charge of approving creative quickly, because slow approvals are the most common reason a good plan produces a thin month.
If you would rather work through the same decision for a piece of software than a service, the software buying process guide walks the same ground: define the requirement, run a real trial, and read the contract before the renewal reads you.
FAQ
How many agencies should I talk to? Two or three. Enough to compare scope, term, and price on the same criteria without stretching the decision over months. Beyond that the proposals start to blur, and the difference between them stops being informative.
Is a specialist agency better than a generalist? It depends on your problem. A specialist that works only with restaurants and cafes will usually need less explaining and will recognise seasonal and weeknight patterns faster. A capable generalist with strong local-search experience can be a better fit if your problem is discovery rather than promotion. State the criteria, then judge each on them.
Should I sign a twelve-month contract for a better rate? Only if you have already seen the work. A discount for a long lock-in is priced for the agency's certainty, not yours. Month-to-month terms are common enough that you can ask for one, and some agencies publish that as their standing policy.
How soon should I expect to see anything? Sooner for engagement and enquiries than for footfall — but ask the agency to put its own expectation in writing rather than accepting an industry rule of thumb. What matters is that you both agreed the review window and the measure before the first invoice, so that month two is a conversation and not an argument.
What is the single most important thing to check? That every account and asset is in the restaurant's name. Everything else — scope, term, reporting — can be renegotiated. Accounts you do not own are the one thing you cannot take with you.
The bottom line
Choosing a marketing agency for a restaurant or cafe is a method, not a leap of faith. Name the problem in one sentence, weight the criteria that fit it, ask two or three agencies the same questions, and compare what a month actually contains rather than what it costs. If you want to see what a restaurant-only agency's published plans and terms look like as a reference point while you build your shortlist, ScaleZone lays its services and terms out on its own site — hold them up against your scorecard alongside everyone else's. Do that and you replace a recurring guess with a decision you can defend, which is what you want from anything that renews every month.
Bettaso may earn a commission when you buy through our comparisons — it never changes the scores. If your next decision is a software one, compare the tools on bettaso.com with the same criteria-first method.
