Marketing agency red flags.

Why the usual list is not very useful

Search for agency red flags and you will get a recognisable set: poor communication, cookie cutter strategy, no relevant experience, bad reviews. All true, all nearly impossible to verify in advance, and all describing the pitch rather than the arrangement.

The useful flags are the ones that show up in the paperwork and the plumbing. They are specific, checkable, and they decide what happens to your business when the relationship ends, which is the moment most people find out what they actually agreed to. Here they are, roughly in order of how much they cost you.

1. Your ads run from their ad account

This is the most expensive one and the one those lists mention least.

If your campaigns live inside the agency's Meta Business Manager or Google Ads account rather than your own, then the conversion history, the audience lists, the pixel or dataset learning and the creative library sit in their asset, not yours. Moving an ad account between business accounts after the fact is difficult, and in practice people leave without the history attached to it.

That history has real value. Paid platforms optimise against accumulated signal, so a new account does not perform like the account it replaced, even with identical targeting and identical creative. Eight months of tuning can reset to zero at the point you change supplier, and you will be told the drop is seasonal.

The correct structure is simple and costs nothing to set up: your business creates and owns the ad account, the pixel or dataset, and the payment method, and grants the agency partner access to it. Good agencies ask for this arrangement themselves, because it removes an argument later.

Ask: will the campaigns run in an ad account owned by my business, and will the pixel be mine? Get the answer in writing.

2. You are not told how the fee is calculated

There are three common ways to be charged: a flat monthly fee, a percentage of your ad spend, or a combination. Each is legitimate. Percentage of spend has been standard practice in media buying for decades, and at large budgets the logic holds up.

At small business budgets it introduces something worth naming out loud. If the fee is a percentage of spend, then the agency's income rises when your spend rises, whether or not the return rises with it. Nobody has to be acting in bad faith for that to shape advice over twelve months. The incentive simply points slightly away from you, and it points hardest at exactly the moment you should be questioning whether to keep spending.

The flag is not the model. The flag is not being told which model you are on, or receiving an invoice where the fee and the media spend arrive as a single number. If those two figures are combined, you cannot audit either one.

A related practice: some arrangements buy media and resell it to you with a margin added. That is a real business model used openly by plenty of firms. It becomes a flag only when it is undisclosed and you believe the invoice shows what the platform charged.

Pricing opacity is not unique to media buying. What SEO actually costs covers the same pattern in the service most often sold without a number attached to it.

Ask: is my fee flat, a percentage of spend, or both? Will the invoice separate your fee from the money that goes to the platform?

3. A long minimum term paired with reluctance to discuss leaving

Long contracts have a defensible reason behind them. SEO and content genuinely take months to show anything, and an agency that hires against your account carries a cost if you leave in week six. Asking for commitment is not predatory.

The thing to look at is proportionality, and the exit. A twelve month minimum with ninety days notice is fifteen months of exposure to a decision you made on the strength of a sales meeting. That may still be reasonable for a large retained programme. It is a lot for a first engagement with someone you have not worked with.

What turns a long term into a flag is the reaction when you ask about it. Someone confident in the work will talk you through what happens if it does not work, and will often offer a shorter initial period, a break clause, or a staged start. Someone who treats the question as a trust issue is telling you the term is doing work that the results are not.

Ask: what is the minimum term, what is the notice period, and what happens if we are three months in and it is not working?

4. The person who sold it to you is not the person doing it

You were pitched by someone senior, experienced and persuasive. That is their job. The question nobody asks is who opens your account on a Tuesday morning.

Junior people doing the work is not a flag. It is how every skilled trade has ever trained anyone, and a well supervised junior with a good process will outperform a distracted senior. The flag is paying senior rates for junior delivery while believing you bought the senior, and never finding out because the only person you ever speak to is the one who sold it.

The second half of this one is caseload. A capable person responsible for twenty accounts is not able to think about yours, no matter how good they are.

Ask: who will do the work day to day, how many other accounts do they hold, and will I speak to them directly or only through an account manager?

5. The report leads with impressions

Here is a test that works on any agency, in any channel, at any budget.

Ask what the monthly report would look like if the campaign had failed.

If there is no version of that report that could show failure, it is not a report. It is reassurance with charts on it. Reach, impressions, follower growth and engagement rate all go up as a near automatic consequence of spending money, which is what makes them comfortable to report and nearly useless for deciding whether to continue.

A real report leads with the outcome you are buying: enquiries, bookings, qualified leads, revenue, cost per acquisition. It states what did not work this month. It is shorter than the other kind.

If you want to be able to read your own numbers rather than take the report on faith, setting up conversion tracking properly is the part to own yourself.

Ask: what will the report lead with, and can I see an example from a month that went badly?

6. Your accounts and assets are in their name

Each of these is small on its own and immovable when you want to leave.

• The domain registered to the agency rather than your business.
• The analytics property created under their account, with you as a viewer.
• The website on their hosting with no way to export it.
• Your Google Business Profile claimed under their login.
• Creative files and raw photography that were never handed over.

The principle is one line: you own everything, you grant them access. If any part of your marketing requires their permission for you to reach it, that is not a service arrangement, and it will be discussed at the worst possible moment.

Ask: will my business hold the admin or owner role on the domain, the analytics property, the hosting, the ad accounts and the Google Business Profile?

If you would rather check before you ask, the free what you actually own tool walks through all eight assets, tells you how to verify each one yourself, and writes the request for the ones that turn out not to be yours.

7. A guarantee

Guaranteed first page. Guaranteed number of leads per month. Guaranteed rankings in ninety days.

Nobody controls Google's ranking, so a ranking guarantee can only be one of three things. It is meaningless, because there is no remedy attached beyond continuing to work for you. It is defined loosely enough to always pay out, which usually means ranking for a phrase nobody searches, including your own business name. Or it is being funded by tactics that carry a risk you were not told about.

Confident people make commitments about their own behaviour instead: what will be delivered, by when, how often you will hear from them, and what they will do if the plan is not working. Those are commitments worth having, because they can actually be kept.

8. They cannot tell you what they would stop

This one is a positive test rather than a warning sign, and it separates strategy from an invoice faster than anything else.

Ask what they would stop doing.

Every real strategy contains subtractions. Fewer channels, fewer posts, one offer instead of five, a page removed, a campaign switched off. Someone who proposes adding nine new activities and removing nothing has not assessed your marketing. They have described their service list back to you.

The answer also tells you whether they looked. "I would stop the Pinterest posting and put that hour into the email list" is the sound of someone who opened your accounts. "We will optimise across all channels" is the sound of someone who did not.

The fair version of all of this

Most agencies are not running a scheme. Long minimum terms, percentage based fees, junior delivery and account managers all have real justifications, and good firms use all four without anyone being worse off for it.

What separates a good relationship from an expensive one is almost never the model. It is whether the model was explained before you agreed to it. Read back through the eight items above and you will notice that none of them are flagged by the practice itself. They are flagged by the answer being evasive, or by the question never being invited.

Which also means you have more control here than the usual advice suggests. You do not need to assess someone's technical skill in a first meeting, which you cannot do anyway. You need to ask eight direct questions and listen to how they are handled.

The eight questions, in one place

1. Will the campaigns run in an ad account owned by my business, with my pixel?
2. Is my fee flat, a percentage of ad spend, or both, and will the invoice separate the two?
3. What is the minimum term and the notice period?
4. What happens if we are three months in and it is not working?
5. Who does the work day to day, and how many other accounts do they hold?
6. What will the monthly report lead with, and can I see one from a bad month?
7. Will my business hold the owner role on the domain, analytics, hosting, ad accounts and Google Business Profile?
8. Looking at my marketing now, what would you stop?

Fair warning: these apply to independent consultants too, including me. If you are weighing up the two models rather than vetting one firm, marketing consultant versus agency covers where each genuinely wins, and the pricing page has my own numbers on it rather than a request to book a call to find out.

If you are already in one of these

Most people reading this are not choosing an agency. They are six months into one and trying to work out whether their unease is justified.

Work in this order. Secure ownership of your accounts first, before any conversation about leaving, because that is the thing that gets difficult once notice is given. Then ask for one report built around enquiries and cost per enquiry rather than reach. A good agency will produce it, and the number may well vindicate them. Then read your contract for the notice period and diarise the date you would need to act by, so the decision stays yours rather than renewing by default.

Only then decide. Changing supplier has a real cost, and a mediocre arrangement you understand is sometimes better than a new one chosen in frustration.