Here is the short answer to how much to spend on Facebook ads while you test. A budget of $75 to $100 a day tells you whether a message lands. It does not tell you whether the ads pay. Knowing whether they pay takes about $2,500 to $5,000 spent and 20 to 50 conversions.
A conversion is the action you are paying for. For most local service owners, that is a form fill or a booked appointment.
Most owners mix those two questions up. They spend $100 a day for a week, see a few leads or none, and decide Facebook works or does not. That week answered a smaller question than they think.
Below, we break the budget into its parts. What each spend level can prove, how to read the first few days, where new ads go, and when to start lower than you planned.
What $75 to $100 a day can tell you
What it is. At this level, you are running a message test. You are asking a simple question. Does anyone stop scrolling, click and fill out the form? And which message pulls harder than the others?
What it costs you when you misread it. Owners treat a message test like a profit test. They crown a winner off three leads. Or they shut the whole channel off after a quiet week. Both calls are made on too little data, and both are expensive.
What good looks like. You read the first stretch as a message test only. You compare ads to each other, not to a profit target. You walk away knowing that one angle gets attention and another does not. That is worth knowing. It is just not the same as knowing the ads pay.
What it takes to know if your ads pay
What it is. To know whether the ads make you money, you need enough results that one good or bad day cannot swing the math. Our working rule is about $2,500 to $5,000 spent and 20 to 50 conversions.
What it costs you when it breaks. With only a handful of conversions, luck looks like skill. Two bookings in a row from one ad can make it look like a star. Two quiet days can make it look dead. Owners who scale or kill on that much data end up guessing with real money.
What good looks like. You pick the right conversion before you start. We judge ads on cost per qualified booked appointment. That is the ad spend divided by the number of real appointments with people who fit. We do not judge on cost per lead, the price you pay for each form fill. We do not judge on CPM, the cost of 1,000 ad views, or on click-through rate, the share of people who clicked. Those numbers help you diagnose. They do not tell you if the ads pay.
We wrote about why the booked number beats the lead number in the math behind closed revenue, not booked calls. Tracking it well also matters, which we cover in why your Meta ads cost per lead can mislead you.
The first 72 hours: what to leave alone
What it is. The first three days of a new campaign are the noisiest. A zero-lead first day is normal on a new pixel. The pixel is the bit of code on your site that tells Meta who took action. On day one, it has not learned anything yet.
What it costs you when it breaks. The owner checks the account on day one, sees no leads, and starts turning things off. Or one ad takes most of the spend, and the owner pauses it to "be fair" to the others. Either way, the money already spent gets thrown away before it could teach you anything.
What good looks like. We act at 72 hours and about $300 spent, not before. One ad taking most of the spend early is the auction picking a winner. We let it. The only reasons to pause a single ad early are defects:
- A wrong link.
- A broken creative, such as a video that will not play or an image cut off.
- A policy problem.
Everything else waits for 72 hours and real spend.
Why one ad on a small budget tells you almost nothing
What it is. At low spend, each single ad holds very few results. When leads cost about $20 each, an ad that has spent $30 to $75 holds 1 to 4 leads. One extra lead can cut its cost per lead in half.
What it costs you when it breaks. You kill good ads on bad luck. A rule like "no lead after a certain spend means kill it" sounds disciplined. At low spend, it mostly measures chance. You end up cutting ads that would have done fine and keeping ads that got lucky.
What good looks like. At low spend, judge groups of ads, not single ads. Group the ads by the idea they share. For example, put every ad that leads with the finished result in one group. Put every ad that leads with the customer's problem in another. Then compare the groups. A group has enough spend and enough leads to say something real.
Where new ads go at each budget level
What it is. Once you have a winner, you will want to test new ads against it. Where you put them depends on your daily budget. An ad set is the group of ads inside a campaign that shares one budget and one audience.
What it costs you when it breaks. The common mistake is building a separate test ad set on a small budget. Below about $100 a day, a separate test set starves both the test and the winner. Neither one gets enough money to tell you anything.
What good looks like. Here is the placement rule we follow:
- Under $100 a day: new ads go into the winner's ad set. We judge them over a full week.
- At $100 a day or more: we clone the winner's ad set into a test set and move about 10% of the budget over. The total budget stays flat.
- At $500 a day or more: each new ad gets its own test set.
When to start lower than $75 a day
What it is. Sometimes the right test budget is smaller on purpose. If your team cannot handle a flood of leads, start at $50 to $75 a day.
What it costs you when it breaks. Cheap leads move the bottleneck to people. The ads work, the forms fill up, and nobody calls them back in time. Then the leads look bad, when the real problem was the follow-up. A lead nobody answers fast is close to worthless, which is why we hold a five-minute line on how fast to respond to a new lead.
What good looks like. You match the budget to the number of leads your team can actually work well. You raise it once replies are fast and steady.
Settings that waste a test budget
What it is. Placements are the places your ad can show up. Some default placements are a poor fit for a small test.
What it costs you when it breaks. Your test money gets spread across places your buyer is not really paying attention. The test takes longer and says less.
What good looks like. These are our build defaults:
- Skip Audience Network. That is the placement that shows your ad inside other apps and sites, outside Facebook and Instagram.
- Skip Marketplace. That is Facebook's buy-and-sell section.
- Load five headlines and five primary texts per ad. The primary text is the main block of words above the image or video. More options give Meta more to work with for each ad.
How the parts fit together
Put the parts in order and you get a simple plan.
- Start with a message test. Run $75 to $100 a day, or $50 to $75 if your team needs a slower start. Use ads that are truly different from each other.
- Leave the first 72 hours alone. Fix defects only. Act at 72 hours and about $300.
- Judge groups, not single ads. At low spend, compare the ideas, not the individual ads.
- Keep going until the answer means something. Look for about $2,500 to $5,000 spent and 20 to 50 conversions before you decide the ads do or do not pay.
- Place new ads by budget. Under $100 a day, into the winner's ad set. Above it, a cloned test set with about 10% of the budget.
Here is how the math plays out. Say you spend $100 a day for 30 days. That is $3,000, inside the range where the answer starts to mean something. If those 30 days produced only a few booked appointments, you still have a message test, not a profit verdict. Keep going, or change the message, but do not call the channel dead.
FAQ
Is $100 a day enough to test Facebook ads?
It is enough to find out whether a message lands. It is not enough, over a week or two, to know whether the ads pay. For that, we look for about $2,500 to $5,000 spent and 20 to 50 conversions. Plan for both stages from the start.
Should you turn off Facebook ads that get no leads on day one?
No, unless something is broken. A zero-lead first day is normal on a new pixel. We act at 72 hours and about $300 spent. Before that, we only pause for a wrong link, a broken creative or a policy problem.
Why is one ad getting almost all of the budget?
That is the auction picking a winner early, and it is normal. Meta pushes the budget toward the ad it expects to do best. Leave it alone unless that ad has a defect. Judge it after 72 hours and real spend.
Should new ads get their own ad set?
It depends on your budget. Under $100 a day, put new ads into the winner's ad set and judge them over a full week. At $100 a day or more, a cloned test set with about 10% of the budget works. At $500 a day or more, each new ad can get its own test set.
If you want a second set of eyes on what your current test budget can and cannot prove, book a discovery call with our team.