For most subscription apps advertising on Meta, an agency starts to pay for itself somewhere between £3,000 and £10,000 a month in ad spend. Below about £3,000, the budget cannot fund enough creative tests on a meaningful event for anyone to learn from, and the fee becomes a third or more of what you spend in total. Published agency minimums run from £3,000 on our own pricing page to the $10,000 to $30,000 that Admiral Media's 2026 pricing guide describes as typical for specialist user acquisition agencies. The better test than either number is your own: how many new ad concepts your budget can read each month at your cost per trial. If the answer is fewer than about four, spend more before you hire anyone.
This guide is about the money. Whether your funnel, tracking and team are ready for outside help is a separate question, and our guide to when a mobile app should hire a Meta Ads agency covers it with a nine-signal framework. Here we set out the formula that turns a budget into a number of readable tests, what the threshold looks like at different costs per trial, what the whole first quarter really costs, and a five-band ladder showing the best value help at each level of spend, including the bands where that help is not an agency.
Key takeaways
- An agency's minimum spend tells you who it will take on. It does not tell you whether your budget can learn anything.
- The useful threshold is readable tests per month. Our rule of thumb is that you need roughly 200 times your cost per trial in monthly spend to read about four new concepts a month.
- The cost per event you read on matters more than the headline budget. £5,000 is plenty at £8 per trial and thin at £60 per subscription.
- Price the first quarter, not the first month. At our minimum spend and starting fee, that is £13,500 before you can judge the result.
- Below about £3,000 a month, a founder, a freelancer or a one-off consultant is usually better value than any agency, ours included.
What agencies publish as their minimum spend
Few agencies publish a minimum at all. Of those that do, the range is wide. We work with a £3,000 minimum monthly ad spend, stated on our pricing page. Admiral Media's guide to user acquisition agency pricing says most specialist UA agencies require a minimum monthly ad spend of $10,000 to $30,000, and adds that some boutique agencies will accept lower budgets for a higher percentage-of-spend fee or a project-based arrangement. RocketShip HQ's consulting and advisory programme is aimed at teams that are past product-market fit, spending at least $10,000 a month and on track to spend at least $100,000 a month within six to twelve months.
Treat all of these, ours included, as sellers' numbers. They reflect the cost of running each agency's service and the clients it wants, which is a reasonable thing for an agency to set but not the same as the point at which your money is better spent with one. We compared what UA agencies charge and why their minimums differ in our guide to mobile app user acquisition agency pricing. The rest of this guide is about the number that is yours.
The real threshold: how many creative tests your budget can read
Most of what a Meta Ads agency does for an app comes down to one loop: make new concepts, put enough money behind each one to tell a winner from noise, read the result and brief the next round from it. If your budget cannot fund that loop, an agency has very little to work with, however good it is. So the question to answer first is how many concepts a month your budget can read.
Our guide to the Meta Ads testing budget sets the two inputs. Around 15% of spend goes to testing once an account has proven winners to carry the rest, and each new concept needs roughly twenty to thirty conversions before it can be judged. For a subscription app, the conversion worth counting is usually the trial start, because installs are where cheap, low-intent traffic hides. Put together:
Cost to read one concept ≈ 25 × your cost per trial start
Readable concepts per month ≈ (monthly spend × testing share) ÷ cost to read one concept
The table applies it at a hypothetical £15 per trial start, so about £375 to read one concept. It shows two testing shares. The steady 15% is right for an account with winners to protect. In the first months, before anything has proven itself, most of the budget is testing in practice, so we also show half. The cost per trial is an illustrative round number, not a benchmark and not client data.
| Monthly Meta spend | Readable concepts at 15% testing | Readable concepts at 50% testing |
|---|---|---|
| £1,500 | Less than 1 | 2 |
| £3,000 | 1 | 4 |
| £5,000 | 2 | 6 |
| £10,000 | 4 | 13 |
| £20,000 | 8 | 26 |
Why four concepts a month as the line? It is one new idea a week. In our experience, set out in that testing guide, health and fitness creative typically fatigues in three to four weeks at scale. An account reading fewer than one new concept a week in a category like that is replacing winners more slowly than they wear out, so its results drift down whoever is running it. At £1,500 a month, even with half the budget on tests, you read two concepts. At £3,000 you read four. That is the arithmetic behind our own £3,000 minimum, and it is why we tell apps below it to put the money into spend first.
The same logic says something about what to buy once you do hire. A creative package of thirty ads a month on a budget that can read four concepts is mostly inventory. Variations of a proven concept are cheap to judge and worth having, but the number of genuinely new ideas you can learn from is set by the budget, not the production quote. We work through that trade in our guide to performance creative agency pricing and in creative diversity versus volume.
Your cost per event moves the threshold more than anything
A spend figure on its own means little, because the same budget reads very different numbers of tests depending on what each event costs you. Rearranging the formula gives the spend needed to read four new concepts a month with half the budget on testing:
Monthly spend to read 4 concepts ≈ 4 × 25 × cost per event ÷ 0.5 = 200 × cost per event
The costs below are hypothetical round numbers chosen to show the range, not benchmarks.
| Event you read tests on | Cost to read one concept | Spend to read 4 a month |
|---|---|---|
| £8 per trial start | £200 | about £1,600 |
| £15 per trial start | £375 | about £3,000 |
| £30 per trial start | £750 | about £6,000 |
| £60 per paid subscription | £1,500 | about £12,000 |
An app in a cheap market with a short, low-friction onboarding can reach a useful testing cadence on well under £3,000 a month. An app that needs to read creative on paid subscriptions, because its trial is long or its trial starts are a poor guide to who pays, needs closer to £12,000. That is also why the threshold usually rises for subscription apps on iOS. The events are deeper, and Apple's attribution postbacks arrive on a delay, so each test takes longer to read, as we explain in our guide to SKAdNetwork postback delays.
This is separate from Meta's own learning threshold. Meta's Business Help Centre says an ad set usually leaves the learning phase after about 50 results in the week after its last significant edit. That decides which event your campaigns can optimise for, and we costed it by event in the hiring-timing guide. The testing threshold here decides how fast you can learn which ads to run. A budget needs to clear both.
Price the first quarter, not the first month
Agency conversations tend to start with the monthly fee. The number that matters is the cost of getting to a reliable answer, and that takes longer than a month. Our pricing page says most accounts see a meaningful signal within 30 to 60 days and a compounding effect from month three, and our engagements start with a three-month commitment for that reason. Other agencies set different terms, so ask, but plan on a quarter whoever you hire.
Using our published starting fee of £1,500 a month for accounts up to £20,000 in spend, the first quarter costs:
- At £3,000 a month in spend: £4,500 a month, £13,500 for the quarter, of which £4,500 is fee.
- At £5,000 a month: £6,500 a month, £19,500 for the quarter.
- At £10,000 a month: £11,500 a month, £34,500 for the quarter.
Now set that against the business. Take two hypothetical apps. One has raised a seed round, has £600,000 in the bank and burns £40,000 a month. A £13,500 quarter is about 2% of its cash. The other is bootstrapped with £40,000 in reserve, and the same quarter is a third of everything it has. The first can afford a quarter whose main output is a clear answer about which messages sell. The second probably cannot, and should spend less, run the account itself and buy help in smaller, cheaper pieces until revenue funds a bigger budget.
A useful test: if the first quarter produced no profitable scale, only a clear picture of what works and what does not, would the business be glad it spent the money? If yes, the budget is ready. If losing that sum would force a decision about the company, you are early. That test also depends on knowing what a subscriber is worth to you, which our guide to setting a target CAC works through.
The spend ladder: the right help at each budget
Put the testing arithmetic, the fee share and the quarter cost together and a ladder emerges. The concept counts use the same hypothetical £15 cost per trial as the table above, so scale them to your own numbers.
| Monthly Meta spend | What the budget can read | Best-value help | Is an agency worth it? |
|---|---|---|---|
| Under £1,500 | Fewer than 2 concepts a month, mostly on installs. | The founder runs the account and makes the first ads. A one-off tracking and account audit is worth paying for. | Not yet. Even our starting fee would be at least half of total outlay. |
| £1,500 to £3,000 | 2 to 4 concepts a month if half the budget goes on testing. | Founder or a freelance media buyer, with creator content bought per video. A consultant for set-up and a second opinion. | Rarely. Most specialist agencies, us included, will not take the account on yet. |
| £3,000 to £10,000 | 4 to 13 concepts a month in the early, test-heavy phase. | A creative-and-media agency if making enough good ads is the constraint. A freelancer if only the media buying is missing. | Yes, if creative supply is the bottleneck. The fee is a third of outlay at the bottom of the band and an eighth at the top. |
| £10,000 to £50,000 | 4 to 20 concepts a month at a steady 15% testing share. | An agency, or a first in-house UA hire paired with a creative partner. | This is where most published specialist UA agency minimums sit, and where an agency is easiest to justify. |
| £50,000 and above | 20 or more concepts a month. | An in-house UA lead with a specialist creative partner, or a full agency team. | Often as a creative partner alongside in-house staff rather than a replacement for them. |
Under £3,000: buy help in pieces
Below our minimum, the cheapest useful help is rarely a retainer. A one-off audit from an experienced consultant can check that installs, trial starts and purchases are reaching Meta and that the account is set up to optimise for the right event, which is the work most likely to waste money if it is wrong. A freelance media buyer can run the account for less than an agency charges, because the job at this level is narrow. Creative can be bought per piece: a few creator videos, a few founder-made ads, and statics built from the reviews and support emails you already have.
The founder's own involvement is an asset here, not a cost to eliminate. At this budget the most valuable thing the account produces is knowledge about which promise makes a stranger install and which objection stops them paying. That is worth learning first hand before handing it to anyone.
£3,000 to £10,000: hire for the bottleneck you have
This band is where the decision is genuinely close, and the answer depends on what is short. If the app already has good ads and the problem is campaign structure and budget management, a freelance media buyer is cheaper and usually enough. If the problem is that a couple of ads carry the account and nobody can make replacements fast enough, that is a creative problem, and an agency that makes the ads and runs the media in one team fits it. At £3,000 our fee is a third of total outlay. At £10,000 it is about an eighth. The fee has to buy back more than that in better ads or founder time, and near the bottom of the band it needs to be a deliberate choice rather than a default.
£10,000 to £50,000: the agency sweet spot
Most specialist agencies' published minimums fall in this band, and for good reason. The budget reads several new concepts a month even at a steady 15% testing share, the fee is a small part of total outlay, and the account is big enough for creative fatigue to cost real money every week it goes unaddressed. The alternative at this level is usually a first in-house UA hire, which works best paired with a creative partner, because one person cannot research, make, launch and read enough distinct concepts on their own.
£50,000 and above: agency, in-house or both
At this spend the question changes from whether to hire help to how to split the work. Our costing of an agency versus an in-house creative team puts the crossover for percentage-of-spend agency fees against a three-person in-house team at roughly £65,000 to £130,000 of monthly spend, depending on the percentage. Above that, many apps keep media buying in-house and use an agency for creative range.
When a bigger budget is not the answer either
Everything above assumes the money has somewhere to go. If users who arrive without an ad rarely start trials or pay, more spend and an agency together will only measure the leak faster, and our hiring-timing guide sets out the signals for that in full. The spend threshold is necessary, not sufficient. A budget that can read four concepts a month is wasted on a funnel that cannot convert the winners. It is worth checking your install-to-trial and trial-to-paid rates against our install-to-trial benchmarks before you scale either the budget or the team, and our subscription app marketing strategy guide shows where paid acquisition sits in the whole funnel.
Where we draw our own line
Our minimum is £3,000 a month in Meta spend because that is roughly where an account at a typical subscription-app cost per trial can read one new concept a week. Below it, we tell apps to put the money into the auction and come back when the account can learn from what it runs. Above it, our Meta Ads service combines creative strategy, production and campaign management for a flat fee rather than a percentage of spend, and our creative refresh calculator will estimate how many new concepts your account needs each month before you speak to anyone.
If your app is spending enough to read tests and the constraint is making enough good ones, see how we work with consumer and subscription apps, look through our case studies, and apply to work with us.
Frequently asked questions
What is the minimum ad spend to hire a mobile app marketing agency?
It depends on the agency, and the published figures vary widely. Our own minimum is £3,000 a month in ad spend. Admiral Media's 2026 pricing guide says most specialist user acquisition agencies require $10,000 to $30,000 a month. An agency's minimum tells you who it will take on, not whether your budget is ready, so check how many creative tests your budget can read before you rely on either number.
Is £1,000 a month enough to run Meta Ads for an app?
It is enough to learn things: which messages get installs, whether tracking works and roughly what an install costs. It is not enough to read creative tests on a trial or purchase event, and an agency fee would be a larger share of the total than the media itself. At that level the founder should run the account and put the money into the auction.
Should I increase ad spend or hire an agency first?
If you are below the point where your budget can read about one new concept a week on the event you care about, increase spend first. An agency cannot learn from tests the budget cannot fund. Once the budget can read tests and the constraint becomes making enough good creative or reading results properly, that is when an agency's fee buys something the extra spend alone would not.
How long should I budget for when hiring an app marketing agency?
Budget for at least a quarter of media plus fees. Our own engagements start with a three-month commitment because research, launch and a proper read take that long, and our pricing page says most accounts see a meaningful signal within 30 to 60 days. At our minimum spend and starting fee, that is £13,500 over three months. If losing that sum with only a clear answer to show for it would hurt the business, you are early.
Does the spend threshold change for subscription apps on iOS?
Usually it goes up. Subscription apps care about trial starts and purchases, which cost far more per event than installs, so the same budget reads fewer tests. On iOS, Apple's attribution postbacks also arrive on a delay, so each test takes longer to read. Both push the point at which an agency can do useful work higher than it would be for an app optimising for installs.
The short version
Before you hire an agency, check that your budget can read about one new ad concept a week on the event you care about. As a rule of thumb that takes around 200 times your cost per trial in monthly spend, which for many subscription apps lands near £3,000 and for apps reading on purchases can be four times that. Then price a full quarter of media and fees against what the business can afford to spend on learning. Below the line, a founder, a freelancer or a one-off consultant is better value. Above it, hire for whichever bottleneck you actually have.