The minimum ad spend worth managing

Ask an agency for its minimum and you usually get a number with no reason attached. The number is real. The reason is more useful than the number, because the reason also tells you what to do while you are below it.

Why do agencies have a minimum ad spend?

Two reasons, and only one of them is about the agency.

The self-interested one first, so it is out of the way. Managing a small account costs an agency almost as much as managing a large one. The same weekly review, the same creative production, the same reporting. If the fee that a small budget can support does not cover a senior person's time, the account is either loss-making or quietly handed to someone junior. Floors are how a shop avoids doing the second thing while claiming the first.

The reason that is actually about you is the platform. Meta's delivery system needs data to learn from, and below a certain volume there is not enough of it for anyone's skill to matter.

What is the learning-phase arithmetic?

Meta's own published guidance on the learning phase is specific: an ad set needs roughly fifty optimisation events in a rolling seven-day window before delivery stabilises and the ad set exits learning. Only the event you optimised for counts — if you optimise for purchases, adds to basket do not help.

Work that backwards and the floor appears on its own. Fifty purchases a week is two hundred a month, per ad set. If your ads convert visitors to buyers at two per cent, that is ten thousand landing page visits a month. LocaliQ and WordStream's benchmark analysis of 1,180 United States campaigns run between April 2024 and June 2025 puts the average Facebook cost per click at $0.70 on traffic campaigns and $1.92 on lead campaigns. Ten thousand visits is somewhere between $7,000 and $19,000 of media, for one ad set, to reach the point where the platform is optimising properly rather than guessing.

The conversion rates in that table are inputs, not measured results — put your own in. What the table shows is not a target. It shows why the answer to “what is the minimum” is never one number: it moves with how well your page converts and how expensive your audience is. A brand converting at five per cent on cheap clicks is properly manageable at a budget that would be hopeless for a brand converting at one per cent on expensive ones.

There is a softer version of the threshold worth knowing. You do not need fifty events per ad set to run ads at all — plenty of accounts spend usefully below it. You need it before optimisation, testing and the whole apparatus of professional management start producing signal rather than noise. Below it, an expensive operator and a careful amateur produce statistically similar results, and only one of them sends an invoice.

So what is the practical floor?

Around $5,000 a month of media is where most of the arithmetic starts working, and there is a second reason for that number beyond the learning phase. Stackmatix's 2026 guide puts market fees at 10 to 20 per cent of spend, or flat retainers from roughly $1,500 upward. At $5,000 of media a $1,500 fee is already 30 per cent of the total. Below that, you are paying more to manage the money than the money can plausibly return.

Add the cost trend and the case gets sharper. The same LocaliQ and WordStream analysis found the average Facebook cost per lead at $27.66, up 20.94 per cent year on year. Small budgets get squeezed hardest by rising costs, because they have the least room to absorb a worse month before the whole test becomes unreadable.

What should you do below the floor?

Spend the money on the two things that make the eventual media work better, both of which you can buy in small amounts.

1. Creative. Pay someone to make ten genuinely different pieces rather than paying someone to manage a budget that cannot test them. Creative is the input with the largest remaining effect on Meta, and it does not expire.
2. The page. Improving conversion from one and a half per cent to three per cent halves every future cost per customer, permanently, and costs less than a month of management.
3. Run the ads yourself, simply. One campaign, broad targeting, your best creative, optimised for the event you actually want. Do not build a structure you do not have the volume to fill.
4. Collect email addresses from every visitor who does not buy. At small budgets, owned audience is worth more per pound than any targeting decision.
5. Write down your margin and your break-even return before you scale. The number you will need later is easier to establish now, on small money.

If you want an outside pair of hands at this stage, hire a freelancer for a fixed piece of work rather than a retainer. A one-off account audit and a creative brief costs a fraction of a monthly fee and gives you most of the value a small account can absorb.

What should you do just above it?

The mistake at the bottom of the manageable range is building the account you will need at ten times the budget. Six campaigns, twenty ad sets, tight audience splits — all of it dividing your fifty weekly events into fragments that each learn nothing.

• Consolidate. Fewer ad sets, more budget each, so events accumulate somewhere rather than everywhere.
• Test creative, not audiences. Creative differences are large and legible at low volume. Audience differences at this budget are mostly noise wearing a label.
• Optimise for the deepest event you can actually hit fifty of. If purchases are too rare, optimise one step earlier and watch the real number behind it.
• Give a change two full weeks. Reading a three-day result at this spend is reading weather, not climate.
• Increase budget in steps rather than jumps, so you do not reset learning on the day you finally got out of it.

And be honest about what management is buying at this level. It is buying discipline and creative volume, not exotic platform technique. Any agency selling you sophistication at $6,000 a month of media is selling you the wrong thing, expensively.

When is hiring us the wrong call?

Below the floor, plainly. If you are spending $2,000 a month, our fee is a larger line than your media and we will tell you so rather than take it. The same is true if your budget is above the floor but seasonal — an account that spends heavily for two months and nothing for ten is better served by a project engagement than a retainer that idles.

There is a third case worth naming because it is the one people argue with. If you are just above the floor and the account is already well run, the honest gain from switching agencies is small. Moving an account costs you a learning reset and a month of ramp. If your current arrangement is boring and works, boring and works is a good outcome.

Floors are not a judgement about your business. They are a statement about when money starts being able to buy an improvement, and it is a better conversation to have before an invoice than after one.