The most common question I hear from small business owners running their first ads: "How much should I be spending?" It's the wrong question, but it's understandable. When you're not sure what to expect, a number feels like something concrete to hold onto.
The right question is: "What am I willing to pay to acquire one customer?" Once you know that, the budget question answers itself.
The Only Math That Matters
Before you set a budget, you need three numbers:
- Your average customer value (ACV): How much does a typical customer spend with you? For a restaurant, it might be $35. For a plumber, it might be $350. For an accountant, it might be $1,500 a year.
- Your target cost per acquisition (CPA): The most you're willing to pay to get one new customer. A reasonable starting point is 20-30% of ACV. If your customer is worth $350, you can afford to spend $70-105 to acquire them.
- Your close rate: What percentage of leads become customers? If you close 50% of people who book a call, and your target CPA is $100, then you can spend $50 per lead.
Average customer value: $400 • Target CPA: $80 (20%) • Close rate: 40% • Max cost per lead: $32 • Monthly budget for 30 leads: ~$960
That's how you set a budget: backwards from what a customer is worth, not forwards from what you're comfortable spending.
What to Spend When You're Just Starting
If you've never run paid ads before, here's the honest truth: your first $300-500 is research money, not revenue money. You're buying data to understand what your audience responds to -- which creative, which offer, which audience segment.
Minimum viable test budget: $15-20 per day, run for 10-14 days. That's $150-280 for a real learning cycle. Any less than that and Meta's algorithm doesn't have enough data to optimize, and you don't have enough data to make decisions.
What you're looking for after two weeks:
- Click-through rate (CTR) above 1%: your creative is stopping the scroll
- Cost per lead under your max threshold: your offer is compelling enough
- Actual leads or sales: the traffic converts on your site
If any of those three are failing, fix that problem before increasing budget. More spend on a broken campaign is just faster money loss.
Realistic Benchmarks by Business Type
These are rough benchmarks based on typical performance for SMBs in competitive U.S. markets. Your numbers will vary based on location, offer, and creative quality.
Local Service Businesses (plumbing, HVAC, cleaning, landscaping)
- Good monthly budget: $500-$1,500
- Expected cost per lead: $15-$50
- Expected cost per booked job: $50-$150
Restaurants and Retail
- Good monthly budget: $300-$800
- Expected cost per click to reservation/location: $0.50-$3
- Focus on local radius, radius under 10 miles
Professional Services (lawyers, accountants, consultants)
- Good monthly budget: $800-$3,000
- Expected cost per lead: $30-$100
- Higher ACV justifies higher CPA -- if a client is worth $5,000, paying $200 per lead is still a good deal
Coaches and Online Services
- Good monthly budget: $500-$2,000
- Expected cost per booked call: $30-$80
- Close rate matters most -- optimize the sales call, not just the ad
When and How to Scale
The worst thing you can do is scale too early. Most business owners see a few leads come in and immediately triple their budget, then wonder why results drop. Here's why that happens: when you dramatically increase a campaign budget, Meta has to re-enter a learning phase. That means higher costs and lower performance for 1-2 weeks.
The right way to scale: increase budget by 20-25% every 5-7 days, not more. This keeps the algorithm stable while letting you grow.
Only scale a campaign when your cost per acquisition has been stable or improving for 7+ days. If it's still fluctuating, you're not ready to scale -- you're still in the learning phase.
How to Allocate Your Budget
If you have $1,000/month to spend on ads, here's a sensible allocation to start:
- 60% Cold traffic campaigns: Finding new customers who don't know you
- 30% Retargeting: People who've visited your site or engaged with your content
- 10% Experimentation: Testing a new creative, new audience, or new offer
As your retargeting audiences grow (more site visitors, more video views), you can shift more budget toward retargeting -- it almost always converts at a lower cost than cold traffic.
The "I Don't Have Budget" Fallacy
One last thing worth saying: "I can't afford to advertise" is usually a math problem, not a money problem.
If your customer is worth $1,000 and you can acquire them for $200, that's a 5x return. The question isn't whether you can afford to advertise -- it's whether you can afford to not advertise. If the math works, borrowing $500 to test a campaign is a reasonable bet.
The business owners who say they can't afford advertising often can't afford to miss the customers who would have found them through that advertising. Run the math first.
The Short Version
Start with $15-20/day and a clear offer. Run for 10-14 days without touching anything. Evaluate cost per lead against your target. Fix the thing that's not working before adding more money. Scale 20-25% at a time. Retarget everyone who shows interest.
That's the whole framework. The budget question matters a lot less than the question of whether the underlying math -- offer, audience, creative -- actually works.
If you want help getting the creative right, start with the Video Script Framework or the Hook Bank Template in the Skills Library.