Most small businesses should start Facebook ads at about $1,000 to $3,000 a month, or $35 to $100 a day, for one campaign. The right number comes from working backwards: what a customer is worth, what you can pay per lead or sale, and how many results Meta needs to optimize, which is about 50 per ad set per week.
Key takeaways
For most small businesses, a realistic starting Facebook ads budget is about $1,000 to $3,000 a month for one campaign, or roughly $35 to $100 a day. You can spend less, but you will get too few results each week to tell what is working.
The exact number comes from three inputs: what a customer is worth to you, what a lead or sale typically costs in your industry, and how many results Meta needs to optimize delivery. A remodeler paying $45 per lead and a T-shirt brand paying $38 per purchase need different budgets, and both differ from a gym paying $25 per lead.
A simple rule: budget for at least 10 results per week so you can make decisions, and for about 50 results per week per ad set if you want Meta's delivery to fully stabilize.
Start with your own numbers, not the ad platform. Work out what you can afford to pay for one customer, turn that into a maximum cost per lead or purchase, then multiply by the number of results you want.
Say the average job is $12,000 at a 35% gross margin, so each customer is worth $4,200 in gross profit. You are willing to spend 25% of that, so your max CAC is $1,050. You close 20% of leads, so you can afford up to $210 per lead ($1,050 x 0.20).
Now the expected cost. LocaliQ’s 2026 Facebook benchmarks put the average lead in Home and Home Improvement at $42.95. Plan on $45. To land 8 jobs a month at a 20% close rate you need 40 leads, and 40 x $45 = $1,800 a month, about $60 a day. Your affordable cost per lead ($210) sits far above the expected cost ($45), so the plan still works if leads run pricier or close rates dip.
If the math only works when leads cost less than your industry average, fix the offer, the price or the sales process before you add ad spend.
Meta's system needs a steady flow of results to learn who to show your ads to. According to Meta’s Business Help Center, an ad set is marked "Learning limited" when it is unlikely to get around 50 optimization events in the week after its last significant edit. An optimization event is whatever you asked Meta to optimize for, such as a lead or a purchase.
That gives you a second formula: weekly budget per ad set = 50 x cost per result. Divide by 7 for a daily number. At $45 per lead that is $2,250 a week, about $321 a day. At $38 per purchase it is $1,900 a week, about $271 a day.
Most small businesses cannot hit that on day one, and that is fine. Learning limited ads still deliver, but results are less stable and often cost more. In an October 2026 best-practice post reported by PPC Land, Meta said ad sets under roughly 50 conversions a week ran about 28% more expensive per purchase in its analysis. Ways to get closer to the threshold:
Costs vary widely by industry, offer, location and season, so treat benchmarks as a starting assumption for your formula, not a promise. Replace them with your own data after 2 to 4 weeks.
Your results can land well above or below these. Strong creative and a clear offer often beat the average, while a slow landing page or slow lead follow-up can double it. One client who reviewed Varsity Agency on Clutch saw cost per B2C lead drop from $65 to $30 to $40 while working with us.
Here are example budgets built with the formulas above. These are illustrations with the math shown, not quotes. The lead and purchase costs are rounded from the benchmarks in the previous section, and the add to cart and ticket costs are assumptions. Swap in your own numbers.
| Business type (example) | Assumed cost per result | Budget for about 10 results a week | Budget for about 50 results a week (learning phase) |
|---|---|---|---|
| Local service lead gen (home services, remodeling) | $45 per lead | 10 x $45 = $450/week, about $64/day | 50 x $45 = $2,250/week, about $321/day |
| Local lead gen with cheaper leads (fitness, real estate) | $25 per lead | 10 x $25 = $250/week, about $36/day | 50 x $25 = $1,250/week, about $179/day |
| Ecommerce, optimizing for purchases | $38 per purchase | 10 x $38 = $380/week, about $54/day | 50 x $38 = $1,900/week, about $271/day |
| New ecommerce store, optimizing for add to cart | $8 per add to cart (assumed) | 10 x $8 = $80/week, about $11/day | 50 x $8 = $400/week, about $57/day |
| Events and ticket sales | $15 per ticket purchase (assumed) | 10 x $15 = $150/week, about $21/day | 50 x $15 = $750/week, about $107/day |
The 10-per-week column is the practical floor for making decisions. The 50-per-week column is where an ad set can fully exit learning. Most local businesses run somewhere in between, and that works. Events are different because spending has a deadline: start early enough to learn which ads sell tickets, then plan for spend to rise as the date gets closer.
In the first month, nearly all of your budget is testing. Once you have proven ads, a simple split is about 70% to 80% on what works and 20% to 30% on testing new creative and offers.
Focus beats spread. Catholic T-Shirt Club went from $0 to $77K in sales in 30 days on one T-shirt with Meta ads. You can see more examples on our results page.
Increase your budget when cost per result has held at or below your target for at least a week and your business can handle more leads or orders. Raise it in steps of about 15% to 25% every 3 to 4 days instead of doubling overnight.
Judge Facebook ads by whether they bring in customers at a cost you can afford relative to what those customers are worth. Platform metrics are clues. Your sales and your bank account are the verdict.
Need help applying this to your account? Our Meta ads team builds budgets this way for both local businesses and ecommerce brands. If you already run ads, book a 30-minute strategy call with Dan and we'll check your budget against your cost per result.
Yes, ads will deliver, but at $5 to $10 a day you may get only a few leads or sales per week. That is too little data to tell good ads from bad ones or for Meta to optimize well. Small budgets work best for retargeting website visitors, promoting a local event or optimizing for a cheaper action like add to cart while you build volume.
Give a new campaign at least 2 to 4 weeks with a budget that produces 10 or more results a week. Judge individual ads after about 7 days, and avoid big changes in the first few days because each significant edit can restart learning. If cost per result is still far above your maximum after a month, change the offer or creative before adding budget.
Use Ads Manager for anything tied to leads or sales. Boosting is quick, but it offers fewer choices for objectives, conversion tracking and testing. Ads Manager lets you optimize for leads or purchases, run several ads in one ad set, exclude existing customers and see cost per result clearly. Boosting is fine for occasional reach on a popular post, not for a lead or sales budget.
For most small accounts, a campaign-level budget (what Meta labels Advantage+ campaign budget) with one or two ad sets is the simplest choice, since Meta moves spend toward what performs. Ad set budgets make sense when you need a guaranteed spend for a specific audience or location, such as a separate service area. Either way, fewer ad sets means each one gets more data.
Usually no. Ad spend is what you pay Meta directly, and management fees, creative production and tools are separate. When you work backwards from customer acquisition cost, include all of them so you see the true cost of a customer. A campaign that looks profitable on ad spend alone can lose money once fees are added, especially at small budgets.
Book a 30-minute strategy call with Dan and we'll show you where your ad budget is leaking and the one fix that moves the needle most.
Book your strategy call →