There’s no single right number, but there is a sensible way to find yours. Start with what a new customer is worth to you, estimate what Meta will charge to win one, and set a budget big enough for Meta’s system to learn from, which usually means enough for about 50 results a week. Rules based on a share of revenue are a useful check, not a target.
Is there a rule of thumb for how much to spend on marketing?
The most common rule is a percentage of revenue. But even the U.S. Small Business Administration doesn’t give one number. In a 2019 article, the SBA wrote that “there’s no hard and fast answer to how much your marketing budget should be,” adding that many businesses use a share of revenue (or projected revenue, for a new business) as a guide, and that the range is wide.
You’ll often read that “the SBA recommends 7–8% of revenue.” We couldn’t find that figure on any SBA page. Its marketing and sales guide only tells you to include a complete breakdown of your marketing costs in your plan and keep tracking them. So treat 7–8% as an often-repeated rule without a clear source.
What do the big surveys say?
Two long-running surveys track how much companies spend on marketing as a share of revenue. They ask different people, so their numbers differ.
- The CMO Survey. In its 2026 report, The CMO Survey (sponsored by Deloitte, Duke’s Fuqua School of Business and the American Marketing Association) found marketing budgets were 9.0% of company revenue, down from 9.4% in 2025. It asked 308 marketing leaders at U.S. for-profit companies in January 2026.
- Gartner. Gartner’s 2026 CMO Spend Survey, published in May 2026, put marketing budgets at 7.8% of company revenue, up slightly from 7.7% in 2025. It asked 401 marketing leaders in North America, the UK and Europe.
Gartner’s numbers come mostly from big companies. In its 2024 survey, the vast majority of respondents reported median annual revenue of over $5.3 billion. That same report showed how much the share has moved: 11.0% in 2020, 6.4% in 2021, then 9.5%, 9.1% and 7.7% in the three years after.
2019 to 2024 from Gartner’s 2024 survey; 2025 and 2026 from its later surveys. Mostly large companies in North America and Europe. Source: Gartner CMO Spend Survey, 2024 to 2026
Do smaller businesses spend a bigger share?
Usually, yes. In the 2026 CMO Survey breakout report, companies with under $10 million in revenue said marketing was 13.3% of revenue on average, and those with $10–25 million said 17.4%. Companies with $1–9.9 billion in revenue said 5.7%.
Take those small-company numbers with care. Each group has only 15 to 34 companies, and answers vary a lot. The pattern makes sense, though: a smaller business has to work harder to be found, and some marketing costs don’t shrink with size.
Averages from 15 to 34 U.S. companies per group; answers vary widely. Source: The CMO Survey, 2026 firm and industry breakout report
How much of a marketing budget goes on ads?
A marketing budget is more than ads. It also pays for people, tools, design, agencies, your website and events. In Gartner’s 2025 survey, paid media was 31% of marketing budgets, Marketing Brew reported. The CMO Survey’s 2026 report found social media took 14.3% of marketing budgets.
Here’s a quick check, with made-up numbers to show the sum. A business turning over $400,000 a year that puts 5% into marketing has $20,000 a year, or about $1,670 a month. If a third of that goes on ads, that’s about $550 a month, or around $18 a day. That’s a ceiling to keep in mind, not a target.
How does Meta decide what your budget buys?
Meta sells ad space through an auction. According to Meta’s Business Help Centre, an auction runs each time there’s a chance to show someone an ad, and billions run every day.
The ad auction is how Meta picks which ad a person sees: the ad with the highest total value wins, not simply the highest bid. That total value combines three things:
- Your bid, what you’re willing to pay for the result you want.
- Estimated action rates, how likely this person is to click, book or buy from this ad.
- Ad quality, judged partly from feedback from people viewing or hiding the ad, and from signs like clickbait or withheld information.
Meta says an ad that’s more relevant to a person can win against ads with higher bids. In plain terms, a better ad costs less per result. That’s why the work you put into the ad itself, especially the first seconds that stop the scroll, matters as much as the size of your budget.
What is Meta’s learning phase, and why does it matter for small budgets?
The learning phase is the period after you launch or significantly change an ad set, when Meta’s system is still working out who to show it to. Results swing more during this time and usually cost more.
Meta says ad sets leave the learning phase once they can deliver stably, which usually happens after about 50 results in the week after the last significant edit. Meta also says a very small budget gives its system a poor signal of who to optimize for.
If an ad set is unlikely to get about 50 results that week, Ads Manager marks it Learning limited. Meta’s page on learning limited says it isn’t a penalty, and lists the usual causes: a small audience, a low budget, a low bid, too many ads at once, or a result that doesn’t happen often. Its fixes are to combine ad sets, widen the audience, raise the budget, or choose a result that happens more often, such as add to cart instead of purchase.
What counts as a significant edit?
Some changes send an ad set back into learning. Meta lists any change to targeting, creative or the result you optimize for, adding a new ad, changing bid strategy, and pausing for 7 days or longer. Budget changes count only when they’re big: Meta’s own example says going from $100 to $101 is unlikely to restart learning, while going from $100 to $1,000 may.
So what’s the smallest useful budget?
A simple floor follows from the 50-result rule: daily budget ≈ 50 × your expected cost per result ÷ 7. If you expect to pay $10 per result, that’s about $71 a day. If you can’t afford that, optimize for something that happens more often (link clicks or landing page views rather than purchases), or accept that the ad set will stay learning limited and judge it over a longer stretch.
What is the minimum daily budget on Meta?
Meta doesn’t publish one fixed minimum anymore. Its best practices for minimum budgets say it sets the minimum based on things like your business type, budget type, bid strategy, what you optimize for, currency and schedule, and that Ads Manager will warn you if you’re below it.
Older articles quote minimums of $1 or $5 a day. Those figures aren’t on Meta’s current page, so check what Ads Manager shows for your own account. One rule Meta does give: if you set a cost per result goal, your daily budget should be at least 5 times that goal, so a $5 goal needs at least $25 a day.
Should you use a daily or a lifetime budget?
Meta’s page on budgets explains the difference:
- A daily budget is the average you want to spend each day, measured over a week. It isn’t a hard cap. Meta is gradually letting some accounts spend up to 75% over the daily budget on days with better opportunities, and less on others, but never more than 7 times the daily budget in a calendar week (Sunday to Saturday).
- A lifetime budget is the total you’re willing to spend over the whole run. It is a hard cap, and Meta paces spending across the dates you set.
Use a daily budget for ads that run all the time, like an always-on offer for a salon or café. Use a lifetime budget when the offer has an end date, like a sale weekend or an event. Either way, Meta says you can also set a campaign or account spending limit so you never spend more than you planned.
How do you work backwards from a goal?
Percent-of-revenue rules tell you what others spend. Working backwards tells you what your ads need to earn. You need four numbers.
- 1Pick one resultBookings, messages, sales or visits: the one thing that tells you an ad worked.
- 2Know what a customer is worthUse the profit from their first year, not just the first sale.
- 3Estimate cost per resultUse your own past ads, or an industry benchmark to begin.
- 4Do the sumCustomers you want times cost per customer gives your monthly budget.
- 5Check it can learnAim for about 50 results a week, or optimize for a more common result.
- 6Run it, then decideGive it two to four weeks before you raise, change or stop it.
- How many new customers you want a month. Be realistic about how many you can serve.
- What a customer is worth. Use the profit from their first year, not just the first sale. A dog groomer’s client who comes back every six weeks is worth far more than one visit.
- What a result costs. Use your own past ads if you have them. If not, start with a benchmark. In its 2026 Facebook ads benchmarks, LocaliQ found the average cost per lead for lead campaigns was $27.39, ranging from $12.30 for career and employment to $61.56 for dentists. Our Meta ad benchmarks guide has more by industry.
- How many results become customers. If one in three enquiries books, your cost per customer is three times your cost per lead.
Cost per customer is what you pay in ads, on average, to win one new customer. Your monthly budget is then the number of customers you want times your cost per customer. If that cost is more than a customer is worth, the plan doesn’t work at any budget: fix the ad, the offer or the page first.
A worked example (illustrative numbers)
Here’s how that looks for a made-up salon. None of these numbers are benchmarks; swap in your own.
| Step | Number | Where it comes from |
|---|---|---|
| New regular clients wanted a month | 10 | The salon’s goal |
| Profit from one regular in year one | $168 | $70 a visit × 6 visits × 40% kept after costs |
| Cost per enquiry | $40 | Assumed |
| Enquiries that become regulars | 1 in 3 | Assumed |
| Cost per new regular | $120 | $40 × 3 |
| Monthly budget | $1,200 | 10 × $120 |
| Daily budget | About $40 | $1,200 ÷ 30 |
Each new regular costs $120 and brings in $168 of profit in year one, so the plan works, with some room for error.
Now check it against the learning phase. Fifty enquiries a week at $40 each would be $2,000 a week, far more than $40 a day buys. So this salon has three sensible choices: optimize for something more common, like visits to its booking page; run one broad ad set rather than several narrow ones; or accept learning limited and judge results on cost per booking over three to four weeks.
What should you start with?

The table below is our practical suggestion for a first test, not data from a study. It assumes one ad set, two or three ads, and an audience that covers the area you serve.
| Your goal | Optimize for | Start with | Give it |
|---|---|---|---|
| Get known nearby | Reach | $5–$10 a day | 2 weeks |
| Visits to your site or booking page | Landing page views | $10–$20 a day | 2–3 weeks |
| Messages or enquiries | Messages or leads | $20–$40 a day | 3–4 weeks |
| Online sales | Purchases, or add to cart at first | $30–$50 a day | 3–4 weeks |
| A dated offer or event | Reach or landing page views, with a lifetime budget | A set total for the run | The offer’s dates |
If you sell online, the purchase line is the one most likely to stay learning limited at small budgets, so many online shops start with add to cart and switch once sales come in.
When should you spend more, and when should you stop?
Decide this before you launch, so the first few days don’t talk you into anything.
Spend more when
- The ad set is out of learning and has been for a week or more.
- Your cost per customer has stayed below what a customer is worth for at least two weeks.
- You can handle more customers.
Raise the budget in steps rather than jumps. Meta’s own example shows small changes rarely restart learning while a tenfold jump may. Our habit is to go up by about a fifth every few days and watch cost per result.
Hold steady when
- The ad set is still learning. Meta says results in this phase aren’t necessarily a guide to later results.
- It’s been only a few days. Costs often swing early on.
Change or stop when
- After three to four weeks, your cost per customer is clearly above what a customer is worth.
- The page people land on is the problem. Check it works on a phone and matches the ad before you blame the ad.
- The same people have seen the ad many times and clicks are falling. That’s creative fatigue, and the fix is new ads, not more money. Our guide to creative fatigue covers how often to refresh.
One catch: pausing an ad set for 7 days or longer sends it back into learning when you switch it on again, so a short break is cheaper than a long one.
What about the cost of making the ads?
There are two costs in advertising, and it helps to keep them apart.
- Making the ads is mostly a fixed cost: a designer, a photographer, an agency, or your own evenings.
- Running the ads is the money you pay Meta, and it grows with your budget.
On a small budget, the making can cost more than the running, so count it. If you spend $600 a month on Meta and pay per ad for design, the design bill can easily match it. And because ads wear out, you’ll need fresh ones every few weeks, not once.
AdForThat’s plans are monthly: Starter is $39 a month, Pro $99 and Ultra $199 (see pricing). Your ad spend is separate. It goes straight from your own Meta ad account to Meta and never passes through us.
Where AdForThat fits
AdForThat makes finished ads for your business from your website, a sentence about what you sell, your logo or a product photo, with no brief and no editor needed. That includes break-outs, where something leaps out of the post and over the feed in your own name, words and colours. In our own experience they’ve been the best ads we’ve ever run for real businesses, with better engagement, more clicks and more conversions than anything else we’ve tried.
Every ad comes in the sizes feeds, Stories and Reels use, and launches to Instagram and Facebook from your own ad account, with your budget set by you. Autopilot keeps your best ads running and puts fresh ones in, up to a monthly spend you choose. See how it works.
Questions people ask
How much should a small business spend on Facebook and Instagram ads per month?
Work it out from your goal: the number of new customers you want times what each one costs you in ads, as long as a customer is worth more than that. Then check the budget can buy about 50 results a week per ad set, which Meta says is usually what it takes to leave the learning phase. Many small businesses start a first test at roughly $10 to $40 a day and adjust from real results.
What is the minimum budget for Meta ads?
Meta no longer publishes one fixed minimum. Its Business Help Centre says the minimum depends on things like your objective, bid strategy, currency and location, and Ads Manager warns you if your budget is too low. If you set a cost per result goal, Meta says your daily budget should be at least 5 times that goal.
What percentage of revenue should a small business spend on marketing?
There is no fixed rule. The CMO Survey found U.S. companies spent an average of 9.0% of revenue on marketing in 2026, and companies with under $10 million in revenue reported 13.3%, from a small sample. That covers all marketing, not just ads, so use it as a check rather than a target.
Does the SBA recommend spending 7 to 8% of revenue on marketing?
We could not find that figure on any SBA page. A 2019 SBA article says there is no hard and fast answer and that many businesses use a percentage of revenue as a guide, with a wide range. The SBA’s business guide only advises including a full breakdown of marketing costs in your plan.
Should I use a daily or lifetime budget on Facebook ads?
Use a daily budget for ads that run all the time, and a lifetime budget for offers or events with an end date. Meta treats a daily budget as an average over a week, so some days may run over, but a week never goes past 7 times the daily budget. A lifetime budget is a hard cap over the whole run.
Why does my ad set say learning limited?
Meta shows learning limited when an ad set is unlikely to get about 50 optimization events in the week after its last significant edit. Common causes are a low budget, a small audience, too many ads at once, or a result that rarely happens. Meta suggests combining ad sets, widening the audience, raising the budget or optimizing for a more common result.
Should I spend more on making ads or on running them?
Keep both in view. Making ads is mostly a fixed cost and running them grows with your budget, so on a small budget the making can cost as much as the running. Ads also wear out, so plan for fresh ones every few weeks rather than a one-off spend.
Sources
- How to Get the Most From Your Marketing Budget, U.S. Small Business Administration (2019)
- Marketing and sales (business guide), U.S. Small Business Administration
- The CMO Survey: Highlights and Insights Report, 2026, The CMO Survey (Duke Fuqua, Deloitte, American Marketing Association) (2026)
- Gartner 2026 CMO Spend Survey finds CMOs allocate 15.3% of marketing budgets to AI, Gartner (press release, via SalesTechStar) (2026)
- Gartner CMO Survey reveals marketing budgets have dropped to 7.7% of overall company revenue in 2024, Gartner (2024)
- The CMO Survey: Firm and Industry Breakout Report, 2026, The CMO Survey (Duke Fuqua, Deloitte, American Marketing Association) (2026)
- Marketing budgets stagnate from 2024 to 2025: report, Marketing Brew (2025)
- About ad auctions, Meta Business Help Centre
- About the learning phase, Meta Business Help Centre
- About learning limited, Meta Business Help Centre
- Significant edits and learning phase, Meta Business Help Centre
- Best practices for minimum budgets, Meta Business Help Centre
- About budgets, Meta Business Help Centre
- How much it costs to advertise on Meta technologies, Meta Business Help Centre
- Facebook Ads Benchmarks for Your Industry, LocaliQ (2026)
Make an ad that does all this
From your website, a product photo or a sentence. Try it free; pick a plan to download and launch.


