How much to invest in Meta ads: how to calculate the budget from your company's numbers

“Let’s try 500 euros a month and see how it goes.” It’s the phrase I hear most often when a company decides to start with Facebook and Instagram. After two months the verdict is almost always the same: Meta doesn’t work for us. In many cases that isn’t true. With that budget the algorithm never had enough data to work out whom to show the ads to, and nobody had decided beforehand which result would count as a success.

The short answer is this: a Meta ads budget is calculated from three numbers. The first is how much you can spend to get a customer or a lead without losing margin. The second is how many results the algorithm needs to learn: according to Meta, around 50 a week for each ad set. The third is how many sales or enquiries you need to reach your commercial goal. If the budget that comes out of the calculation is higher than you can afford, you don’t cut at random: you reduce the number of ad sets, choose a more frequent optimisation event or give Meta a different role.

Below you’ll find the method we use at the agency, with two complete examples (an e-commerce store and a B2B company) and the Meta rules worth knowing before you put a figure down.

Where do you start when deciding a Meta ads budget?

You start from the financial result you want to achieve. The budget is the consequence of two numbers: how many sales or enquiries you need and how much you can pay for each one. A figure decided in advance (“a thousand euros a month”) is only useful as a ceiling, to be compared with the calculation.

There are two ways of reasoning, and it’s worth using both:

• top-down: what share of turnover or of the overall marketing budget you want to put into advertising on Meta. It’s the percentage method, useful for understanding what the company can sustain. I discuss it in the article on the B2B marketing budget;
• bottom-up: how many results you need and how much each one costs. This is the method that tells you whether that figure is enough to make the campaigns work.

When the two numbers are close, you’re fine. When the bottom-up number is much higher than the top-down one, you have a problem to solve before spending the first euro.

How do you calculate how much you can spend on a customer or a lead?

You calculate it from the margin left on each sale. The maximum cost per acquisition is the part of the margin you’re willing to give up to advertising to win a customer. Beyond that threshold, every sale costs you more than it leaves you.

For an e-commerce store the calculation is direct. For a B2B company, which sells off the website, you need an extra step: from the value of a customer you work out the value of a lead, using the salesperson’s close rate.

Step	E-commerce (example)	B2B company (example)
Value of a sale	Average order €80	First year of contract €12,000
Margin after variable costs	€32 (product, shipping, returns)	€4,800 (40%)
Share of margin allocated to acquisition	75%: €24	25%: €1,200
Conversion rate to customer	-	1 lead in 20 becomes a customer (5%)
Maximum cost per result	€24 per purchase	€60 per lead

These numbers are examples: yours need to come from your management system and CRM. Two words of caution. In e-commerce, if the customer comes back to buy again, the margin to consider is that of the first 6-12 months, but only if you’ve already measured repeat purchases. In B2B the close rate should be read on leads coming from campaigns, which usually close less often than those that arrive through word of mouth.

If you don’t yet have these numbers in your CRM, the first job is to get them: without them, any budget is a gamble. How to connect campaigns to opportunities and revenue is something I explain in the article from lead to contract.

How much budget do you need to exit the learning phase?

You need enough budget to get around 50 results in the week after the last significant edit, for each ad set. Meta says so on its page about the learning phase: below that threshold performance doesn’t stabilise and the cost per result generally stays higher.

The sum is simple: minimum weekly budget = 50 × expected cost per result.

	E-commerce (purchases)	B2B (leads)
Expected cost per result	€24	€60
Weekly budget for 50 results	€1,200	€3,000
Equivalent daily budget	about €170	about €430
Equivalent monthly budget	about €5,200	about €13,000

For ads with checkout inside Meta’s shop the rule is different: you need at least 17 purchases on the website and 5 through Meta within seven days.

If the ad set doesn’t get there, Ads Manager shows the status Learning limited. Meta points out on its page about learning limited that this isn’t a penalty: it means the budget isn’t being spent effectively with that setup. The causes listed are almost always the same: small audience, low budget, cost cap too tight, rare optimisation event, too many ad sets running at once.

And if the budget that comes out of the calculation is too high?

It’s the most common situation for an SME, especially in B2B. There are four levers, to be used in this order.

1. Fewer ad sets. The 50 results apply to each ad set. With five ad sets sharing 300 euros a day in total, each one sees a fifth of the data. Meta itself recommends combining similar ad sets, because combining them also combines what they learn. For an SME one campaign with one or two ad sets is often enough.

2. A more frequent optimisation event. If 50 purchases a week are out of reach, you can optimise for add-to-baskets or checkouts initiated, which cost less and happen more often. It’s the same example Meta gives. The price you pay is a signal further from the sale: you need to check that people who add to the basket then go on to buy.

3. A further-off but qualified event, in B2B. For leads the problem is the opposite: form fills are frequent, but often of poor quality. Here it’s worth optimising for the lead and sending back to Meta, through the Conversions API, which leads became qualified. With low volumes the algorithm won’t be able to optimise directly for that event, but it gets an indication of whom to look for.

4. A different role for Meta. If even then the numbers don’t add up, Meta can do another job: making the brand known to decision-makers in a small market, supporting a trade fair, re-engaging people who visited the website. In that case the budget is decided on reach and frequency, and success is measured in brand searches and direct enquiries.

What’s best avoided is splitting a small budget across lots of campaigns “to try everything”. The result is that none of them leaves the learning phase and in the end you haven’t learnt anything.

What’s the minimum budget on Meta ads?

The technical minimum is low, and usually far from the budget that makes a campaign work. Meta states, on its page about best practices for minimum budgets, that minimums depend on industry, budget type, bid strategy, optimisation, currency and country, and that Ads Manager shows a warning when the budget is below the threshold or too low for optimal results.

Three rules from the same page are worth knowing:

• With a cost per result goal, the daily budget must be at least 5 times the goal. If you want leads at 20 euros, the daily budget must be at least 100 euros.
• The event you choose changes the budget you need. Optimising for purchases requires more budget than optimising for landing page views.
• With Advantage+ campaign budget (what used to be called campaign budget optimisation) the risk of underspending is lower, because the budget moves freely between ad sets.

Hence the rule of thumb we give clients: the minimum sensible budget is the one that brings at least one ad set close to 50 results a week. Below that, you’re paying for visibility, which can make sense, but it needs saying up front.

Daily or lifetime budget, at campaign or ad set level?

For most ongoing campaigns a daily budget at campaign level (Advantage+ campaign budget) is best. A lifetime budget makes sense for initiatives with an end date, such as an event or a promotion. An ad set budget is used when you need to guarantee a minimum spend to a specific segment.

Choice	When to use it	What to know
Advantage+ campaign budget	Ongoing campaigns with several ad sets	Meta moves the budget to where it sees most results; you can set minimum and maximum limits per ad set
Ad set budget	When a segment must receive a defined spend (a country, a customer audience)	With budget sharing on, up to 20% of the daily budget can move to the other ad sets
Daily budget	Always-on campaigns	On a single day spend can exceed the budget by up to 75%, but over the week it stays within 7 times the daily budget
Lifetime budget	Events, launches, sales with precise dates	If you reduce it, the new amount can’t go below what’s already been spent plus 10% of the spend from the last two days

The point about the daily budget explains a lot of alarmed Monday phone calls: one day you spend more, another less, but the week evens out. The flexibility and sharing figures are those given by Meta on its page about ad set budget sharing.

How do you split the budget between acquisition, remarketing and testing?

In most cases the biggest share goes to acquiring new people, a small share to remarketing and a fixed share to creative testing. The proportion depends on the size of the audience that already knows you.

In the projects we work on we start from this split, and adjust it after the first month:

Item	E-commerce	B2B	Why
Acquisition (broad audience, Advantage+)	70-80%	60-70%	This is where new customers are found
Remarketing (visitors, unconverted leads, customers)	10-15%	20-25%	In B2B the cycle is long and re-engagement matters more
Creative testing	10-15%	10-15%	Without new ads, performance declines over time

Remarketing has a natural limit: you can only spend in proportion to how many people you have to re-engage. With a few thousand visitors a month, too high a remarketing budget translates into people seeing the same ad ten times. You’ll find how it works on the glossary page about remarketing.

Another Meta rule applies to testing: lots of ads and lots of ad sets fragment learning. A few tests at a time, each with a precise question (a video format against a carousel, a message about price against one about delivery), are better than ten variants at once.

When and how should you increase the budget?

You increase it when the ad set has left the learning phase and the cost per result is below the maximum you calculated. You increase it in steps.

Meta explains on its page about the last significant edit that budget changes can restart learning, depending on how big they are. Pausing and changing audience, optimisation event or creative always do. So:

• read the data from the date of the last significant edit, not from the start of the month (the column can be added in Ads Manager);
• increase in moderate steps, one at a time, and wait at least a few days before the next one;
• avoid touching budget, audience and ads in the same week: if things change, you won’t know why;
• if an increase makes the cost per result rise sharply, go back to the previous level instead of changing everything else.

There’s also a market limit. Doubling the budget doesn’t double the results: beyond a certain point Meta has to go looking for less interested people, and each additional result costs more. The signal is a cost per result that rises as frequency increases. It’s time to widen the audience, refresh the creatives or open a new market.

How can you tell whether the budget is working?

By looking at the cost per result against the maximum calculated at the start, and comparing Meta’s results with those in your management system or CRM. The Ads Manager figure alone isn’t enough, because it’s the result Meta attributes to itself.

Since 2026 this matters even more. Meta has changed the way it counts clicks: since the end of March 2026, for campaigns optimised for conversions on the website or in store, only a click on the link counts as a click-through. Likes, shares, saves and videos watched for at least 5 seconds have moved to a separate category. Anyone comparing 2026 with 2025 without knowing this sees a drop that is partly just a change in the rules. Gaetano explains it in the article on misleading ROAS.

The checks we do every month:

1. Cost per result from the date of the last significant edit, compared with the sustainable maximum.
2. Share of new customers out of total attributed sales: if the budget goes mainly to people who already know you, you’re paying for sales you’d have made anyway.
3. Total sales or leads in the management system or CRM, compared with the previous month and with overall spend.
4. Lead quality, for B2B: how many became qualified and turned into opportunities. A lead scoring system helps read them all the same way.
5. Frequency and ad fatigue: when frequency rises and CTR falls, you need a new creative before a bigger budget.

Mistakes to avoid

The costliest mistakes concern how the budget is distributed and judged.

• Judging Meta after two weeks. With 50 results a week as the benchmark, the first two weeks are for the algorithm. Before deciding you need at least four to six weeks without major changes.
• Splitting a small budget across too many campaigns. Each ad set learns on its own. One that works is better than six in learning limited.
• Changing the budget every day. Every significant edit calls into question what the algorithm has learnt.
• Setting a cost per lead without looking at the margin. A lead at 15 euros is expensive if it never closes; one at 90 is good value if one customer in five signs.
• Measuring only inside Ads Manager. The reference point is revenue or pipeline in the CRM. The rest is the platform’s point of view.
• Spending without complete tracking. Without the pixel, Conversions API and consent properly set up, Meta optimises on partial data and the budget delivers less.

Checklist for calculating the budget in seven steps

1. Take the average margin per sale or per contract from your management system or CRM.
2. Decide what share of that margin you can allocate to acquisition and work out the maximum cost per result.
3. In B2B, divide by the close rate of campaign leads to get the maximum cost per lead.
4. Multiply the expected cost per result by 50: that’s the minimum weekly budget per ad set.
5. Compare that number with the budget the company can sustain. If it’s higher, reduce the ad sets, choose a more frequent event or give Meta a different role.
6. Split the budget between acquisition, remarketing and testing, starting from the percentages in the table and adjusting them after the first month.
7. Before launch, set the monthly checks: cost per result, share of new customers, results in the CRM, frequency.

Want to know how much it makes sense to invest in your case?

When a company asks us how much to put into Meta, we start from its numbers: margins, close rate, reachable audience, tracking. We tell you what budget you need to make the campaigns work, what to expect in the first three months and whether, with the numbers you have today, Meta is the right channel or whether it’s better to start from another one. Campaign work is the job of our Meta ads service, with tracking handled by our analytics and data service.

You can book a call with Federico or write to us via the contact form, telling us your industry, average margin and goal for the next six months.

Frequently asked questions

What’s the minimum you need to invest in Meta ads?

Meta’s technical minimum is low and varies by country, currency and objective. The useful minimum is the budget that brings at least one ad set close to 50 results a week: you calculate it by multiplying the expected cost per result by 50. With a cost per lead of 30 euros, that’s about 1,500 euros a week.

Is a daily or a lifetime budget better?

A daily budget for always-on campaigns, a lifetime budget for initiatives with an end date, such as an event or a promotion. With a daily budget Meta can spend up to 75% more on a single day, but over the week it stays within 7 times the daily budget.

How long does the learning phase last on Meta?

Until the ad set gets around 50 results in the week after the last significant edit. With an adequate budget a few days are enough; with too low a budget the ad set stays in “Learning limited”, which isn’t a penalty but a sign that the setup needs changing.

What should you do if the ad set stays in learning limited?

Reduce the number of ad sets and combine similar ones, widen the audience, increase the budget or the cost cap, or choose a more frequent optimisation event. These are the solutions Meta itself recommends.

How often can you increase the budget?

After leaving the learning phase and with a cost per result below the sustainable maximum. It’s best to proceed in moderate steps, one at a time, waiting a few days between increases, because significant budget changes can restart learning.

Do Meta ads work for B2B with small budgets too?

Yes, if you give them the right role. With small budgets and expensive leads it’s hard to reach 50 results a week, so Meta works better as a remarketing and awareness channel towards a defined audience, measured on brand searches and direct enquiries. To generate leads in volume, in B2B it’s often best to pair it with Google Ads and LinkedIn.