How to Choose a Marketing Agency (and Spot the Ones That Can't Deliver)

Seven questions that separate agencies who run a real, accountable system from those who sell you leads and disappear.

The worst agency relationships I've seen didn't fail. That's the part nobody warns you about.

They succeeded, precisely, against the thing that was written down. The agency promised a number of leads a month and delivered that number of leads a month, every month, on time. The reports were accurate. The invoices were accurate. And at the end of it the owner had almost nothing to show for six months and a pile of money, because nobody had ever agreed on what was supposed to happen after a lead arrived.

That is the most common way this goes wrong in aesthetics, and it does not look like failure while it's happening. It looks like everything is working.

So the job of choosing an agency is not really about finding someone competent. Most of them are competent at the thing they actually do. The job is finding out what they are on the hook for, and making sure the part they are on the hook for is the part that turns into money.

Here are the seven questions I'd ask if I were hiring, in the order I'd ask them.

What "done-for-you" is supposed to mean

Before the questions, one piece of vocabulary, because it's the word that causes most of the confusion.

Almost every agency says done-for-you. Very few of them mean the same thing by it, and almost none of them define it before you sign.

For a lot of agencies, done-for-you means they do a task for you. They build the ads. They run the ads. That's a real service and there's nothing dishonest about it. But it leaves you holding every handoff: the lead arrives and you catch it, you answer it, you follow up, you book it, you remind them, you chase the no-show. If you have staff to do that, fine. Most owners don't, which is why they hired an agency in the first place.

Done-for-you should mean the handoffs are owned. Not that somebody somewhere is doing marketing tasks, but that between a stranger seeing an ad and a consultation appearing on your calendar, there is a defined owner for every step, and it isn't you.

That distinction is what the seven questions are really testing.

One: who owns the lead after it arrives?

This is the whole thing, and I'd ask it first, before pricing, before anything.

There is a version of this business where the agency's job ends the moment somebody raises their hand. They send it to you and their part is over. And there's a version where the agency is responsible all the way to a booked, confirmed consultation on your calendar.

Both are legitimate. They are wildly different products and they are frequently sold at similar prices under similar language.

A good answer is specific and it draws a line you can see. Ours ends at a qualified consultation that is booked and confirmed, and the provider owns the consultation itself. That's a boundary. You know exactly where our part stops and yours starts.

A bad answer is enthusiastic and vague. "We handle everything." "We take care of all of it." Push once and ask where their responsibility ends, and if they can't tell you, it ends at the lead.

Two: what happens in the first five minutes after somebody raises their hand?

Not "do you follow up." Everyone says yes to that.

Ask what happens in the first five minutes. Then ask what happens at nine o'clock on a Saturday night, because that's when a lot of people in this industry are actually sitting on their phones deciding what to do with their face.

A real answer has a mechanism in it. Who or what responds, how fast, through which channel, and what happens if that first message doesn't get a reply. A weak answer is an adjective. "We respond quickly." "We're very responsive." Ask what quickly means in minutes and watch what happens.

Then ask the follow-up question, which is how many times they'll reach out before they stop. If the answer is once or twice, or if there is no answer because nobody has ever counted, that tells you the follow-up is a person's good intentions rather than a system. Good intentions do not survive a busy Tuesday.

Three: where do my leads live, and what happens to them if I leave?

Ask where a lead physically goes when it comes in.

If the answer is that it goes to your Instagram inbox, or an email address, or a spreadsheet somebody updates, you do not have a system. You have a pile. And the practical consequence is the thing every owner in this industry has done at least once: scrolling backward through direct messages at ten at night trying to remember who you were talking to and what they wanted.

The answer you want is that every lead from every source lands in one CRM, with a pipeline, where you can see who came in, what they asked about, what's been said to them, and what stage they're in.

Then ask the harder half. Whose account is it? If you fired them tomorrow, do you keep the ad account, the CRM, the lead history, and the phone number? The answer should be yes and it should be boring. If it gets complicated, you are renting your own customer list.

Four: what exactly are you responsible for, and what am I responsible for?

In writing. Both columns.

The second column is the one people skip, and it's where most of the resentment in these relationships comes from a few months in. If the agency needs you to approve creative within a certain window, provide photos of your work, or record what happened at each consultation, and nobody ever said that out loud, both sides end up feeling let down and neither one is wrong.

A good agency will hand you your responsibilities without you asking, and some of them will be things you'd rather not do. That is a good sign, not a bad one. It means they've done this enough times to know where it breaks.

Be suspicious of an agreement that describes only what they do. Every real system has homework.

Five: what will you report, and what can you not report without me?

This one separates people faster than any other question, and it's the one I'd pay the most attention to.

Anyone running paid traffic can report spend, leads, cost per lead, conversations started, and consultations booked. That's their system, so that's their data, and you should expect all of it.

Here's the part almost nobody volunteers: an agency cannot tell you what you actually earned unless you tell them what happened after the consultation. Whether the person showed up. Whether they bought. What they spent. That information exists only in your business, and if you don't put it somewhere the agency can see, nobody on earth can calculate your real return. Not us, not anyone.

So ask what they can report on their own, and ask what they need from you to close the loop. An agency that answers that honestly is telling you they've thought about attribution seriously. An agency that promises full revenue reporting without asking anything of you is either about to make numbers up or about to attribute every sale you were going to make anyway to their ads.

We tell clients plainly that if outcomes don't get recorded, the honest answer at the end of the month is that return on investment cannot be calculated yet. That's an uncomfortable sentence to say to someone who is paying you. It's also the truth, and I'd rather say it than dress up a number.

Six: how long before this works, and what happens if I stop?

Paid traffic has a ramp. When ads first turn on, the platform doesn't know who your buyer is yet, so early leads usually cost more and convert worse while it figures that out. That improves over weeks, and it keeps improving after that. This is not a story agencies tell to buy themselves time. It's how the auction works.

Which means two things you should ask about.

First, ask what month one is supposed to look like versus month three, specifically. Someone who has done this a lot can tell you. Someone who can't is either new or hasn't been paying attention to their own accounts.

Second, ask what happens if you pause. This matters more than people expect. Turning ads off for two weeks doesn't cost you two weeks. It costs you the two weeks plus the ramp you have to climb again when you switch back on, and you gave up the compounding you already paid for. If you're planning a vacation, the better move is almost always to leave it running and let the calendar fill ahead of you.

Seven: what do you not do?

Ask it exactly like that, and then be quiet.

An agency that says it does everything is telling you it does nothing in particular. The good answer is a list, delivered without embarrassment, of things that are genuinely outside the scope. Organic social. Content production. Photography. Clinical anything. Whatever it is.

You want the exclusions before you sign, not in month four when you ask for something and discover it costs extra. It will cost extra. That's fine and it's normal. Finding out about it in advance is the difference between a business relationship and an argument.

The test that isn't a question

Here's the one I'd actually weight most heavily, and it doesn't require you to ask anything.

Watch what they want from you before they start.

An agency that asks for your logo, your ad budget, and access to your Facebook page, then tells you they'll be live Monday, is going to run generic ads. They have to. They don't know anything about you.

An agency that puts you through a genuinely tedious exercise first, your pricing, your services, who you think your competitors are, what offers you've run before and how they did, what makes your work different, is doing the part that decides whether any of it works. Specific ads outperform generic ones, and the platform itself rewards that specificity by showing distinctive ads to more people. You cannot write a specific ad about a business you don't understand.

We put every client through a Brand Context Sheet before we build anything, and I'll be honest that people don't enjoy it. One client resisted it for six months. When she finally sat down and did it, she told me it was the single most useful exercise she'd ever done on her own business, and she was annoyed at herself for waiting. That document changes what we can write, how we position her, and which offers we put in front of which people.

If nobody asks you hard questions before they take your money, they are not planning to do that work.

About commitments

You'll be told to avoid contracts. I don't think that's right, and I'll tell you where I land, including the part where I have skin in it.

We ask for three months up front, so read this knowing that.

A commitment is reasonable when someone can explain exactly what the time is for. Ours is for the ramp: getting the system built, getting through the platform's learning period, and getting far enough into optimization that the numbers mean something. We've had clients hit month three, watch it finally click, and leave anyway, and then come back later and start the whole climb over. That's the thing the commitment is trying to prevent, and it's a real thing, not a sales device.

A commitment is a warning sign when nobody can tell you what happens inside it. If you ask what month one looks like versus month three and get a vague answer, the length isn't protecting your results. It's protecting their revenue.

Ask what the time is for. The answer tells you which kind you're looking at.

Where we are not the right fit

I'd be doing exactly what this article warns about if I wrote seven questions that only we pass. So here are the situations where you should hire someone else, and I mean this literally.

If you want your organic social run, daily posting, blog writing, search engine optimization, a website built, or photography, that is not what we do, and you should hire the person who does. We do produce video for ads, because that is part of running them. We run the paid acquisition system end to end and keep our hands off the rest. An agency that says yes to all of it and to ads is spreading itself thin somewhere.

If you're not willing to record what happened at your consultations, don't hire anyone who talks about return on investment, and that includes us. Not because we'll be annoyed, but because the conversation in month three will be useless. You'll ask whether it's working and the honest answer will be that nobody knows.

And if you can't fund meaningful ad spend on top of a management fee, wait. The fee buys you a system. The system needs fuel. A well-built engine with nothing in the tank produces the same result as no engine, and you'll have paid for the privilege.

One more thing that isn't about us. Nobody honest will guarantee you leads, appointments, sales, or revenue. Not us, not anyone. Too much of the outcome depends on your work, your prices, your reviews, your availability, and your market. If someone offers you a guarantee, you've just learned the most useful thing you're going to learn about them.

What to do with this

Take the seven questions to whoever you're talking to. Ask them plainly. You are not being difficult, you're being a buyer, and anyone who treats the questions as an insult has answered them.

Pay attention less to the answers themselves than to whether they're specific. Vague is the tell. Every one of these questions has a real answer, and people who run real systems give it immediately, because they've had to think about it.

If you'd like us to be one of the agencies you ask, book a call. We'll go through your actual numbers, tell you where your money is currently leaking, and be straight with you about whether we're the right fit. Sometimes we're not, and we'll say so on the call rather than three months into an invoice.

Koru & Ivy does not guarantee leads, appointments, sales, rankings, or revenue. Results depend on your market, offer, pricing, capacity, and execution, and vary between businesses.