Most people pick a budget by gut feel, then judge the ads against it. Start from what a customer is worth to you instead. Put in your numbers and this works out the most you can pay for a lead and a booked appointment, and what to spend each day testing ads.
Want a second pair of eyes on these? If you're spending $7k or more a month on Meta and the numbers don't add up, book a call. If it's a fit, we'll map it out. If it's not, I'll tell you straight.
Book a callA budget built on someone else's cost per lead gives you someone else's budget. This one only uses your prices, your margins and how many of your leads become customers.
These are the real numbers from the solar business I co-own. An average install is $15,000 and costs about $1,250 to deliver. I'll spend up to $2,500 to win a job.
About one in every 8.5 leads becomes a customer, so the most I can pay for a lead is $2,500 divided by 8.5, about $294. In practice, I want it under $150.
About one in three leads books an in-home appointment. That's roughly 2.8 appointments for every customer, so the most I can pay for a booked appointment is about $882.
Testing 4 new ads at $50 a day each is $200 a day, or $800 for each four-day round.
Your budget doesn't decide whether ads work. It decides how many you can test at once. Here's how I use the numbers above on every new ad.
Every new ad gets its own ad set and budget. Don't touch it for four days. Checking daily leads to tinkering, and tinkering stops you reading the result.
At four days, compare its cost per lead with your ceiling above. Over the line, it gets cut. Cost per lead is the fastest number, so it's the early warning.
An ad that passes runs to around 20 leads. Then judge it on how many became qualified, booked appointments. That's the number it lives or dies on.
Ads booking appointments under your ceiling earn more budget, by no more than about 20% in any 24 hours. Keep launching new ads beside them.
The full method, with the Qvolt numbers worked through: How do I calculate my Facebook ads budget?
Enough to give each new ad its own ad set at a proper testing budget, about $50 a day, times the number of ads you want to test at once. Most new advertisers start with 4 ads, which is $200 a day. Your budget decides how many ads you can test, not whether ads work.
Divide the most you'll spend to win a customer by the number of leads it takes to win one. That's your ceiling. For Qvolt Solar it's $2,500 divided by 8.5 leads, about $294. Use it as the kill line when you check a new ad at four days.
Benchmarks tell you what other people paid. They don't tell you what you can afford to pay. They don't know your prices, your margins or how many of your leads become customers, and those three things are the whole calculation.
What a booked appointment costs you. Check cost per lead at four days as an early warning. If the ad passes, let it run to around 20 leads, then judge it on how many became qualified, booked appointments. Sales per ad are too slow and too few to judge an ad on.
No. A bigger budget lets you test more ads at once. It doesn't make a weak ad work. Start with as many ads as you can fund properly, and give more budget to the ones that prove themselves.
Fix that before you set a budget. Put booked and sold stages in your CRM so every lead can be traced past the form. Until you can do that, any budget is a guess.
I build and run Meta ad campaigns for service businesses that care about booked appointments, not a spreadsheet full of cheap leads.
Book a call