Meta ads agency UK: what a good one does in 2026, what to ask, and the red flags

A good Meta ads agency in the UK in 2026 is mainly a creative agency with a media buyer attached. Meta's system now does most of the targeting on its own, so the thing that moves your results is the flow of new ads going into the account, how honestly they are read, and what gets made next because of it. Before you sign, ask how many new ads they will ship each week, how they decide an ad has won or lost, and who owns the account and the creative. If the answers are vague, keep looking.

Most of the pages ranking for this search are lists of agencies, and most of the checklists stop at "ask about tracking" and "watch for hidden fees". Both matter. But they miss the part that decides whether you grow: what the agency does on a normal Tuesday. That is what this article is about, drawn from the accounts we run across the UK, US, GCC and EU.

What a Meta ads agency actually does now

Three or four years ago a Meta agency earned its fee by building audiences: interest stacks, lookalikes, careful exclusions. That work has mostly been absorbed by Meta's own delivery system. Broad targeting with strong creative is now the default, and the ad itself is what tells Meta who to show it to.

So the work has moved. A good agency in 2026 spends its week on four things:

1. Reading the account. Which ads, angles and formats cleared your target last week and which missed, measured against your numbers, not an industry average.
2. Deciding what to make. A written plan for the next batch: who each ad is for, what it says, and what format it takes, based on what the last batch taught.
3. Making the ads. Statics and videos, in volume, in your brand's own look.
4. Buying media around them. Campaign structure, budgets, scaling what works and switching off what does not.

The fourth item is the one most agencies sell. The first three are the ones that produce results. When we look at an account that has stalled, the cause is almost never a bid setting. It is nearly always that the same handful of ads have been running for months and nobody has given Meta anything new to find buyers with.

One example from our own work: on a UK luxury home brand, the ads written for people who already love the category missed the target badly, at under a third of the return we needed across six ads. The ads written for the buyer who hesitates on price cleared it. No targeting setting would have found that. Only testing different messages to different people did. We wrote about the same pattern in the first August Weekly.

The questions to ask before you sign

Use these on the first call. A good agency answers each one with a number or a specific process. A weak one answers with adjectives.

Question	A good answer sounds like	A weak answer sounds like
How many new ads will go live each week?	A number, tied to your spend level	"As many as we need"
Who makes the creative?	A named team or process, with examples in your category	"We work with your in-house team"
How do you decide an ad has won?	A target return or cost per result, and a minimum spend before a call	"We look at the data"
What do you report on?	Cost per purchase or lead, return against your target, what was learned	Reach, impressions, CTR
Who owns the ad account, pixel and creative?	You do. They are added as a partner	Hesitation, or "we run it in ours"
What happens in the first 30 days?	Tracking checked, a first batch live, a first read of results	A long "discovery phase"
Can I speak to a current client?	Yes, and here are published results	"Our clients are confidential" with nothing public

Two of these deserve more time.

How do you decide an ad has won? This is where most agencies are weakest, and it matters because a wrong call kills good ads or scales bad ones. Our view, after a lot of batches, is that a winner is rare and relative. In one account this month, most new ads received under £20 of spend before Meta moved its budget elsewhere. You cannot read those as losers, because they were never really tested. An agency should tell you how much spend an ad needs before they will call it, and they should be willing to say "not enough data yet" out loud.

What do you report on? Ask to see a real report from another client, with names removed. If the headline numbers are reach and clicks, the agency is reporting what is easy, not what pays you. The report you want says what was spent, what came back against your target, which ideas won, which lost, and what is being made next as a result.

How many ads a good agency should ship

There is no single right number, because it depends on spend. But the direction is clear: more new ads than most brands expect.

We run 20 to 50 new ads a week across the accounts we manage. The reason is simple arithmetic. Most new ads will not beat your current best. If one in ten does, you need ten tries to find one, and you need to find one regularly, because every winner fades.

Volume without variety is wasted, though. Ten versions of the same video with different opening lines is one test, not ten. Ask the agency how their batches differ from each other: different buyers, different reasons to buy, different formats. On one ecommerce account we manage, statics have cleared the return target across more than 50 ads over the last six weeks. On another, UGC style video has come in at about two thirds of target across eight. On a premium product, we have seen our best static return more than twice what our best video did in the same week. Those are findings from specific accounts, not rules. The point is that you only learn them if the agency tests both properly and tells you honestly which one won.

For a deeper look at what these patterns look like week to week, the August Weekly shares what our accounts taught us each week.

The red flags

Some of these are covered elsewhere. We have added the ones we see when brands come to us after a bad experience.

• A promised return before they have seen your numbers. Nobody can promise a ROAS without knowing your margins, your price and your site's conversion rate.
• The ad account or pixel is in their name. If you leave, you lose your history. You should own the Business Manager and add the agency as a partner.
• Nobody can tell you who makes the ads. If creative is "outsourced" with no names or examples, it is usually the first thing to slip.
• The same ads have been live for three months. Check this yourself in Meta's Ad Library. Search your brand, look at the start dates.
• Reports that change shape every month. When results are bad, the chosen metrics tend to move. Agree the headline numbers at the start.
• A long discovery phase before anything goes live. You should have new ads running within the first couple of weeks.
• Everything is a winner. If every update is good news, you are not being told the truth. Real testing produces a lot of losers, and a good agency shows them to you.

What results should look like, and how fast

Honest expectations help you judge an agency fairly. In our experience the first month is mostly learning: tracking checked, a first batch or two live, and a first real read on which ideas and formats work for your buyers. By the end of the second or third month you should see the trend moving on the number you care about.

A few examples from our published case studies, where the brands are happy to be named:

• Posh MIA went from a 0.7x ROAS to a consistent 3.2x in three months.
• Liquidation Store scaled Meta spend nearly 4x in a month while ROAS rose from 4.2x to 5.7x.
• Le Luxe grew orders from Meta 3.4x in two months, with ROAS up 24% at the same time.
• Azelefa went from never running ads to a consistent 4x ROAS in two months.

The common thread in each is not a clever setting. It is a steady supply of new ads, read honestly, with the next batch built from what the last one taught.

Do you need a UK agency specifically?

Not necessarily, but UK experience helps in a few places. UK buyers respond to a different tone from US buyers: less hype, more proof. Pricing, delivery promises and returns language need to match UK expectations and the site. And some categories, such as health, finance and anything making a claim about results, sit under UK advertising rules that a US team may not know well.

What matters more than the office address is whether the agency has run accounts like yours, at your spend, selling to your buyers. Ask for that evidence directly.

A quick checklist before you sign

• You own the ad account, pixel, Business Manager and all creative
• They have told you how many new ads go live each week
• They have explained how they call a winner, including the minimum spend
• You have seen a real report and it leads with cost per result and return against target
• New ads will be live within the first two weeks
• You have seen published results or spoken to a current client
• The notice period and every fee are written down

If you would like to see what our approach looks like on your own brand before a call, we will make you 10 free ads to judge us by.

FAQ

How much does a Meta ads agency in the UK cost? Most charge either a fixed monthly fee or a percentage of ad spend, sometimes both. Whichever model you choose, make sure creative production is included or priced clearly, because it is where most of the value sits now.

How long should I give a new agency? Around three months for a fair read. The first month is mostly learning. If there are no new ads live in the first two weeks, you do not need to wait that long to know.

Should the agency make the ads or just run them? In 2026, the agency that makes the ads should be the one reading the results. When creative and media buying are split between two teams, the lessons from one rarely reach the other.

Is a bigger agency better? Not by default. Ask who will actually work on your account each week and how many other accounts they handle.