How much to spend on Meta ads (and when to scale)
A simple budget formula with a worked example, cost-per-lead benchmarks by industry, a five-step plan for scaling and what to do on a small budget.
Start with enough budget for about 50 results a week in each ad set: your target cost per lead × 50 ÷ 7 is a sensible daily budget. At the median Facebook cost per lead of US$27.66 in WordStream’s 2025 data, that is about $198 a day. If you have less, run one campaign with one ad set and judge it over a month. Scale only after your cost per lead has held for a week, and raise the budget in small steps so the ad set doesn’t go back into the learning phase.
- The usual benchmark for leaving Meta’s learning phase is about 50 results in 7 days for each ad set.
- A starting formula: target cost per lead × 50 ÷ 7 = daily budget. It is a rule of thumb, not an official Meta rule.
- WordStream’s 2025 data puts the median Facebook cost per lead at US$27.66, from $16.61 in real estate to $76.71 for dentists.
- Meta may spend up to 75% over your daily budget on a single day, but no more than seven times that budget in a week.
- Scale like you water a plant: a little at a time, with a few days between each increase.
Key terms explained (16)
- Meta ads
- Paid ads on Facebook, Instagram, Messenger and other apps owned by Meta.
- Ads Manager
- Meta’s free tool for building ads, setting budgets and seeing results.
- Campaign, ad set and ad
- The three levels of a Meta ad account. A campaign holds your goal. An ad set holds the budget, audience and schedule. An ad is the picture, video and words people see.
- Learning phase
- The period after you launch or change an ad set, when Meta is still working out who to show it to. Results swing more during this time.
- Learning limited
- The label Ads Manager shows when an ad set isn’t getting enough results to leave the learning phase.
- Optimization event
- The result you ask Meta to get for you, such as a lead, a purchase or a booked call.
- Significant edit
- A change big enough to restart the learning phase, such as a large budget jump, new ads or new targeting.
- Cost per lead (CPL)
- What you spent on ads divided by the number of leads you got. $500 spent for 20 leads is a $25 cost per lead.
- Click-through rate (CTR)
- The share of people who clicked your ad after seeing it.
- Cost per click (CPC)
- What you paid, on average, each time someone clicked your ad.
- Conversion rate
- The share of people who did what you wanted, such as filling in a form, after clicking.
- Daily budget
- The average amount you tell Meta to spend each day. Meta can go over on some days and under on others.
- Creative
- The image, video and words in an ad.
- Median
- The middle value in a list of numbers. Half are higher and half are lower.
- Benchmark
- A typical result from other businesses that you can compare your own numbers against.
- CRM
- Customer relationship management software: one place that stores your contacts, conversations, appointments and deals.
How much should you spend on Meta ads?
Spend enough for Meta to learn who your customers are. Meta ads are the paid posts you run on Facebook and Instagram, and Meta’s system decides who sees them. It gets better at that as results come in. The benchmark most advertisers use is about 50 results in 7 days for each ad set (a group of ads that share one budget and audience).
That gives you a simple formula. Take your target cost per lead, which is the most you’re happy to pay for one enquiry. Multiply it by 50, then divide by 7. This formula is my rule of thumb for a starting point. It isn’t an official Meta rule.
Worked example. In WordStream’s Facebook Ads Benchmarks 2025, the median cost per lead across 726 US lead campaigns was US$27.66. So: $27.66 × 50 = $1,383 a week, and $1,383 ÷ 7 = about $198 a day. A home improvement business, where the median is $41.26, would need about $295 a day.
If those numbers look too high for you, don’t stop reading. Plenty of businesses get good results on less. The small-budget section shows how.
Meta ads budget calculator
A planning estimate, not a promise. Your real cost per lead depends on your offer, your ads and your market.
What is the learning phase?
The learning phase is the period after you launch or change an ad set, when Meta is still testing who to show your ads to. Results and costs swing more during this time.
Meta’s own developer documentation lists three learning statuses: learning (“the ad set is still learning”), success (it has left the learning phase) and fail (it “isn’t generating enough results to exit the learning phase”). In Ads Manager, the tool where you build and manage Meta ads, that last one shows up as “Learning limited”. The same page says “significant edits cause ad sets to reenter the learning phase”.
Where does the number 50 come from? Meta ads specialist Jon Loomer explains in The learning phase is just a label (October 2025) that advertisers work to a benchmark of 50 optimized events in a seven-day period. An optimization event is the result you asked Meta to get you, such as a lead or a purchase.
He also makes a point worth remembering: the label is a guide, not a hard rule. Meta is always learning, and you can get good results while an ad set still says “Learning”. His advice is not to tinker constantly, and not to avoid important changes either.
What counts as a significant edit?
- A big budget change. Small increases are usually fine. Large jumps are not.
- New or edited ads. Changing the image, video or text.
- New targeting. Changing who the ad set is aimed at.
What does a lead cost on Meta ads?
It depends heavily on your industry. Here is the median cost per lead (CPL) for Facebook lead campaigns from WordStream’s 2025 data. A median is the middle value: half of the campaigns paid more and half paid less.
Across all industries in the same report, the median click-through rate (CTR, the share of people who clicked after seeing the ad) was 2.59%. The median cost per click (CPC) was $1.92. The conversion rate, the share of clicks that became leads, was 7.72%, down from 8.67% a year earlier. Cost per lead rose about 21%, from $22.87 to $27.66.
Two cautions. These are US campaigns priced in US dollars, so treat them as a rough guide if you advertise in Canada or elsewhere. And a cheap lead isn’t always a good one. What matters is the cost of a lead who becomes a customer, which is why I track leads through to sales in a CRM (customer relationship management software).
Why does Meta sometimes spend more than your daily budget?
Because a daily budget is an average, not a daily limit. In Updates to Meta ads budgeting, Jon Loomer quotes Meta’s own wording: “we may spend up to 75% over your daily budget” on days when better opportunities are available. Meta won’t spend more than seven times your daily budget across a week.
| Your daily budget | Most Meta may spend in one day | Most Meta will spend in a week |
|---|---|---|
| $20 | $35 | $140 |
| $50 | $87.50 | $350 |
| $100 | $175 | $700 |
| $200 | $350 | $1,400 |
So judge your spend by the week, not by the day. One expensive Tuesday is normal.
When should you scale your Meta ads?
Scale when the ads are working and the rest of your business can keep up. Check all four before you add money:
- Your cost per lead has held at or under target for at least 7 days. One good day isn’t a trend.
- The ad set has left the learning phase, or its results are steady even if the label still says “Learning”.
- The leads are good. They answer the phone, they fit your service area and some of them buy.
- You can handle more leads. If replies already take hours, fix that first with the five CRM automations.
How do you scale without resetting the learning phase?
1. Check that the numbers have held for 7 days
Look at cost per lead over the last 7 days, not yesterday. If it’s at or under your target and lead quality is good, move on.
2. Raise the budget a little, not a lot
A common rule of thumb among advertisers is to raise the budget by about 20% at a time. Meta doesn’t publish a fixed number. Sometimes Ads Manager tells you directly: Jon Loomer shares an example where it said “You can increase your daily budget up to $179 without reentering the learning phase” on a $100 budget. If you see a message like that, use it.
3. Wait 3 to 4 days before you touch it again
This is my working habit, not a Meta rule. Costs often wobble for a day or two after an increase. Give it time to settle before you decide whether the change worked.
4. Add new creative before you add new ad sets
Creative means the image, video and words in your ad. When results flatten, fresh creative usually helps more than new audiences. Splitting your budget across more ad sets makes it harder for each one to reach 50 results. I cover this in Meta ads for local service businesses.
5. Set a ceiling
Decide in advance when you’ll stop: when cost per lead rises above your target for a full week, when lead quality drops, or when your team can’t follow up fast enough. Then step back to the last budget that worked.
What if you have a small budget?
You can still run Meta ads. You just need to make it easy for Meta to learn from fewer results. This table uses the $27.66 median cost per lead to estimate what each budget buys.
| Daily budget | About how many leads a week | What to do |
|---|---|---|
| $10 | 2 to 3 | One campaign, one ad set, two or three ads. Judge it after a month, not a week. |
| $20 | 5 | Same setup. Keep targeting broad so Meta has room to find people. |
| $30 | 7 to 8 | Same setup. Swap in one new ad every two to three weeks. |
| $50 | 12 to 13 | Still one ad set. Enough to compare two offers over a month. |
| $100 | 25 | Halfway to 50 a week. Results should feel steadier. |
| $200 | 50 | Around the 50-a-week benchmark at the median cost. |
Estimates only: daily budget × 7 ÷ $27.66. Your own cost per lead will differ.
Three things help most on a small budget:
- Don’t split it. One ad set with $20 a day learns faster than four with $5 each.
- Choose a result that happens more often. If booked appointments are rare, ask Meta for leads instead, then turn leads into bookings with fast follow-up. My follow-up text scripts are built for this.
- Be patient. With five leads a week, one week tells you very little. Look at a month.
What budget mistakes should you avoid?
- Doubling the budget after one good day. It’s the whole bucket on the plant.
- Editing every day. Each big edit can restart learning. Batch your changes and make them once a week.
- Too many ad sets. More ad sets means fewer results in each one.
- Judging by the day. Meta can spend up to 75% over your daily budget on a single day. Look at the week.
- Counting leads, not customers. Track which leads become sales, and report it monthly with a one-page marketing report.
- No tracking on your website. If you send people to a landing page, set up analytics first. My GA4 setup guide takes about 10 minutes.
Tools for the job
| Tool | Use it for | Cost |
|---|---|---|
| Meta Ads Manager | Building ads, setting budgets and checking the learning status | Free to use; you pay for the ads |
| Google Analytics 4 | Seeing what ad visitors do on your website | Free |
| Google Sheets | Tracking weekly spend, leads and cost per lead | Free |
| GoHighLevel | Following up with leads in minutes and tracking which ones buy | Paid |
Plans and pricing change. Check each tool’s site for current details.
Frequently asked questions
How much should a small business spend on Meta ads?
Use the formula target cost per lead × 50 ÷ 7 as a starting daily budget. At the median Facebook cost per lead of US$27.66 in WordStream’s 2025 data, that is about $198 a day. If you have less, run one campaign with one ad set and judge results over a month.
What is the Meta ads learning phase?
It is the period after you launch or significantly edit an ad set, when Meta is still learning who to show your ads to. The usual benchmark for leaving it is about 50 results in 7 days. Results and costs are less stable until then.
How much can I increase my Meta ads budget at once?
A common rule of thumb is about 20% at a time, with a few days between increases. Meta doesn’t publish a fixed number, and Ads Manager sometimes shows how far you can raise a budget without re-entering the learning phase.
Why did Meta spend more than my daily budget?
A daily budget is an average. Meta may spend up to 75% over it on days with better opportunities, but it won’t spend more than seven times your daily budget in a week.
Can I run Meta ads on $10 a day?
Yes. At the median cost per lead that is two or three leads a week, so keep everything in one campaign and one ad set, and give it a month before you judge it.
What is a good cost per lead on Facebook ads?
It depends on your industry. In WordStream’s 2025 data the median was US$27.66 overall, ranging from $16.61 for real estate to $76.71 for dentists. A good cost per lead is one that leaves you a profit once leads become customers.
Related playbooks
Browse every playbook on paid media, or see all marketing playbooks.
Sources
- Facebook Ads Benchmarks 2025WordStream by LocaliQ
- Ad Campaign Learning Stage InfoMeta for Developers
- The learning phase is just a labelJon Loomer Digital
- Updates to Meta ads budgetingJon Loomer Digital
Remar Catapang is a full-funnel marketing strategist based in Ontario, Canada. He has worked in marketing since 2020 and continues to lead search, CRM automation, web, paid media and experiential campaigns for businesses in Canada, the United States, Australia, Egypt and the Philippines. His playbooks are based on that ongoing work. Questions about this playbook: remarcatapangjr@gmail.com
