Set your monthly ad budget from the number of results you need and the most you can afford to pay for each one: budget equals results needed multiplied by your affordable cost per result. Then check that the figure is big enough for the platform to learn from, and small enough that a weak month does not hurt the business.
Why a percentage of revenue is a starting point, not an answer
A percentage of revenue tells you what the business can afford to spend on marketing in total. It does not tell you what your ads need to buy, so it works as a ceiling, not as a plan.
The best-known figure comes from large companies. In the Gartner CMO Spend Survey, marketing budgets sit at 7.8% of company revenue, and most respondents work for organisations with more than $1 billion in annual revenue.
That average covers the whole marketing budget, not just ads, and it describes businesses many times larger than most advertisers. Use it as a sense check after you have done the sum below.
What a percentage tells you
- How much the business can spend without straining cash
- Whether marketing is in line with similar firms
What it leaves out
- How many leads, sales or bookings you need
- What each result is worth to you
- Whether the platform gets enough data to learn
Work out your budget from the result you need
Start from the result, not the spend. Decide how many leads, sales or bookings you need each month, decide the most you can pay for each one, and multiply.
An online store that needs 100 orders a month and can pay up to $30 per order: 100 × $30 = $3,000 a month
The same sum works for any goal. Only the result and its price change.
| Goal | Affordable cost per result | Results needed a month | Monthly budget |
|---|---|---|---|
| Leads for a service firm | $80 per lead | 40 | $3,200 |
| Online sales | $30 per order | 100 | $3,000 |
| Bookings for a clinic or studio | $45 per booking | 60 | $2,700 |
| Store visits | $6 per visit | 500 | $3,000 |
Four budgets of similar size buy very different numbers of results. That difference matters in the learning check further down.
Monthly ad budget calculator
Enter the results you need and the most you can pay for each one.
Illustrative defaults. Nothing you type is saved or sent anywhere.
The most you can pay per lead, sale or booking
The affordable cost per result comes from your own margins and close rates, not from an industry average. It is the most a result can cost before advertising stops paying for itself, minus the profit you want to keep.
Cost per order
Start from what one order keeps after product, shipping, fees and returns. A $100 order at a 40% contribution margin keeps $40; keep $10 of it as profit and you can pay up to $30.
- Read more
- Break-even ROAS
Cost per lead
Take the gross profit from one customer, the share of it you will spend to win them, and the rate at which leads become customers. The $80 in the table sits under a $120 maximum worked out that way.
- Read more
- A good cost per lead
Cost per booking
Use the lead method, then count no-shows. If one booking in five is a no-show, you can pay a fifth less per booking than per visit.
Tip: use first-order or first-year value when cash is tight, and lifetime value only when repeat business is reliable and you can wait for it.
The minimum a platform needs to learn
A budget also has to buy enough results for the platform to settle. Below that level, delivery stays unstable and costs usually run higher, whatever the sum says.
Meta says an ad set becomes learning limited when it is unlikely to reach about 50 optimisation events in that week, and names low budget as one cause. It also advises a budget “large enough to get enough total results”.
On Google, Target ROAS on Search and Shopping needs at least 15 conversions in the past 30 days before you can use it.
About 218 results a month. At $30 an order that is $6,540 for one ad set; at $80 a lead, $17,440.
If your budget falls short of that, shape the account around it rather than spreading it thinner. Meta lists these options itself:
- Run fewer ad sets, so each one gets more of the results
- Combine similar ad sets and campaigns
- Optimise for an event that happens more often, such as add to basket instead of purchase
- Widen the audience so more people can complete the event
Splitting a first budget between channels
Fund one channel properly before you spread money across several. A budget split four ways can leave every channel short of the data it needs.
- Start where buyers are closest to acting Search catches people already looking for what you sell; social ads reach people who are not looking yet. Lead with the one that matches how your customers buy.
- Run a short, cheap test if you are unsure A small split test across channels can show which one produces results at all, before the main budget is committed.
- Keep a small slice for testing Hold back part of the budget for new ads or a second channel, and judge it against the same cost per result.
- Move money on evidence When one channel brings results under your affordable cost and another does not, shift the budget and say why in writing.
A budget is not a number you pick. It is the price of the results you need, checked against what the platform needs to learn.
When to raise the budget, and when not to
Raise the budget when results are steady below your affordable cost and more volume would be welcome. Hold it when the problem is something money cannot fix.
Raise when
- Cost per result has held under your affordable cost for a few weeks
- Ad sets are out of learning and delivering steadily
- You can handle more orders, leads or bookings
Hold when
- Ad sets are still learning, or you changed them this week
- Cost per result is already above your affordable cost
- The offer, page or close rate is the weak point
Raise in steps rather than all at once. Meta recommends avoiding frequent budget changes, which can send an ad set back into learning, so give each change about a week of results. If performance has dropped, our guide on what to check before you cut budget covers the other side.
Check the budget against your own numbers
In a strategy session we work out the budget with you from your margins and close rates, and the plan is yours to keep.
What real budgets bought in our projects
Three projects show the same logic at different sizes. Each started from a clear result and a fixed amount, not from a percentage.
The brief was a five-times return on $1,200 in seven days. Splitting the budget by product line showed the incense burners, with a third of the money, bringing just over half of the 45 orders.
- Budget
- $1,200 in seven days
- Result
- $6,216 in sales, a 5.18 return
After several phases of moving budget towards the strongest product and audiences, the last phase ran at $8.47 per order, down from $29.65 at launch.
- Budget
- $2,000 in a month
- Result
- 236 orders, about $11,700 in sales
A $400 split test across Facebook, Instagram, Search and YouTube came before any larger spend. Search and YouTube sold nothing, so the plan scaled Instagram.
- Budget
- $400 test
- Result
- $25 per sale on Instagram, $100 on Facebook
In the course project, on one peak sales day the budget was raised to $1,000 while the return was 7.1 and pulled back when it dropped. The budget followed the results, in both directions. That is the habit our media buying work is built on.
If you want the full arithmetic in a spreadsheet, the Ad Budget & Break-even Planner turns your margins into target costs, test budgets and a monthly plan.
In short
- Budget = results needed a month × the most you can pay per result
- Work the cost per result out from your own margins and close rates
- Check the budget can buy about 50 results a week per Meta ad set, or shape the account around it
- Fund one channel properly before splitting
- Raise in steps when results hold under your affordable cost
What owners ask about ad budgets
Should ad spend be a percentage of revenue?
A percentage of revenue is a useful ceiling, not a method. It shows what the business can afford in total, while the results-based sum shows what the ads need to buy. The budget you run should pass both checks.
Is a small budget worth running at all?
Yes, if it answers one clear question rather than trying to run a full account. In one of our projects a $400 split test across four channels showed which one to fund. A small budget spread across many ad sets is much harder to learn from.
Should the agency fee count as part of the ad budget?
No. Plan the ad budget from your cost per result and keep the fee as its own line, then add both together when you judge your true cost per customer. At AdvanSEM, ad spend is separate from the fee and never marked up.
How often should we change the budget?
Change it in steps and give each change about a week of results before the next one. Meta advises against frequent budget changes because they can send an ad set back into the learning phase.
What daily budget should a new campaign start with?
Divide your monthly figure by 30.4, then check that it can buy at least a few results a week at your affordable cost. If it cannot, run one campaign with one ad set, or optimise for an earlier step such as add to basket.