Facebook ads for online courses
Facebook ads for online courses: what they cost and when they pay off
The short answer: Facebook ads don't have a price list. You pay in an auction to reach specific people, so the same campaign costs different amounts depending on the offer, the audience and how good the creative is. The only number that really decides profitability is your cost per acquisition (CPA) set against the margin one sold course leaves you.
In practice, a sensible starting point is around $2,500 a month in ad spend — you can launch with less, but the benefits of scale arrive later. The first signals of whether a campaign has a chance usually show up within 7–14 days. Give it about 60 days for a measurable impact on CPA and ROAS. Below, we break down exactly what drives the cost and how to calculate your own break-even point before you spend your first dollar.
Written by Grzegorz Świerad, founder of Subida AI.
Three numbers to start with
- monthly ad spend — the level where we work best
- ~$2,500monthly ad spend — the level where we work best
- first signals of whether the campaign has a chance
- 7–14 daysfirst signals of whether the campaign has a chance
- measurable impact on CPA and ROAS
- 60 daysmeasurable impact on CPA and ROAS
What drives the cost of Facebook ads?
When someone asks "how much do Facebook ads cost", what they're really asking about is the cost of acquiring a buyer. That cost is made up of several things, and only some of them are on Facebook's side.
The offer and course price
The higher the price, the longer the decision and the more expensive each purchase — but also the bigger the margin you can spend on acquiring a customer. A cheap course sells faster but leaves less room for ad costs.
The creative and the first three seconds
An ad nobody watches takes up as much space in the auction as one that sells. Differences in CTR and in how many people stay past the first shot translate directly into cost per click. That's why producing many variants is cheaper than paying to prop up one weak creative.
Auction competition and seasonality
You're bidding for the same person's attention alongside every other advertiser. At peak season — New Year's resolutions, Black Friday, back-to-school — prices rise because the number of bidders grows, not the number of users.
The funnel and the sales page
Facebook delivers traffic, but it's the page that turns it into orders. The same ad sent to two different sales pages produces two completely different CPAs. Most often, the bottleneck is the page, not the ad.
Conversion data quality
A correctly installed pixel, Conversions API and event deduplication determine what the algorithm learns from. A campaign fed incomplete data optimises toward the wrong goal and burns through budget before anyone notices.
The campaign objective
A free webinar sign-up, a form lead and a course purchase are three different prices. Comparing "cost per lead" across campaigns with different objectives makes no sense — compare the cost of the same event.
What budget does it make sense to start with?
We work best from around $2,500 a month in ad spend. That isn't a pricing policy or an entry threshold designed to screen out smaller clients — it's the amount at which you can keep a campaign alive and test messaging at the same time. You can launch with less, but the benefits of scale arrive later.
The reason is mechanical. Meta's algorithm learns from conversion events: until a campaign collects them regularly, it stays in the learning phase and delivers unstable results. Too small a budget means too few conversions to exit that phase — and at the same time too little money to test more than one messaging idea at once.
The second reason is testing. With a single set of creatives, you can't tell whether a result comes from the product, the audience or chance. That's why we split the budget deliberately: part of it works on sales, part of it buys knowledge you then use in every campaign that follows.
How many days until you know whether a campaign works?
- 01
Days 1–14: first signals
You can see which creatives hold attention, which messages get clicks at all, and whether traffic reaches the checkout. This isn't a financial result yet — it's information about which direction to go.
- 02
After the first two weeks: shaping the campaign
Winning creatives get more budget, losing ones get switched off. We fix the page where people stall. Usually in this window CPA starts settling into a repeatable number instead of jumping around from day to day.
- 03
Around day 60: measurable impact
By day 60 you can usually see a measurable impact on CPA and ROAS — meaning you can compare the campaign's economics with what you had before launch and decide whether to scale or pull back.
One caveat: these timeframes apply to campaigns with working tracking. If conversions aren't flowing back to Meta, none of these timelines hold, because you're measuring noise.
When does an online course ad pay off? Run the numbers yourself
This is the one question nobody can answer for you, but the formula can be laid out. Plug in your own values and you'll get the threshold above which the campaign loses money — instead of guessing whether your CPA is "good" by industry averages.
Step 1. Calculate your unit margin (M)
From the course's gross price, subtract VAT or sales tax, payment processing fees, platform costs and the real cost of serving one buyer — support, materials, any instructor fee. What's left is M: the money you actually have for acquiring that customer and for profit.
Step 2. Set your break-even point (maximum CPA)
Maximum CPA = M. When acquisition cost equals your unit margin, you break even: the campaign pays for itself, but you make nothing. That's the ceiling, not the target.
Step 3. Set your target CPA
Target CPA = M minus the part of the margin you want to keep in the business. If you want to keep half the margin, you aim for half of M. This one decision defines the whole campaign: the budget, the cost per click you can accept, and how aggressively you can scale.
Step 4. Convert it to break-even ROAS
Break-even ROAS = net price divided by M. If your margin is half the net price, the threshold is 2 — below that the campaign loses money. It's the same information as in step 2, just expressed in the unit the ad dashboard shows.
Step 5. Only now add what comes later
If you have a second product, a higher price tier or renewals, calculate M not from a single transaction but from customer value over a chosen time horizon. Then you can pay more to acquire a customer than the margin on their first purchase — provided you can fund that gap until the money comes back.
Two traps that break this calculation
Calculating margin from the gross price. VAT and sales tax were never your revenue, and including them in M can inflate the break-even point so much that the campaign is "profitable" only in the spreadsheet.
Treating ROAS from the ad dashboard as gospel. Ads Manager reports attributed conversions, not booked revenue. The final reckoning is done on bank transfers, not in the dashboard.
Does it work? Three projects with a concrete result
Not all of them are online courses, but the customer-acquisition mechanics were the same: Facebook ads, lots of tested creatives, and a funnel designed for fast iteration.
“Cisza w świecie dziecka”: around 1,000 copies in the first month
A digital product brought to market with an ad-driven launch. Around a thousand copies sold in the first month.
Silne Dziecko (Kamil Kajak): 350 pre-orders
A custom funnel and store designed for rapid experiments. The pre-sale alone closed at 350 orders, which came to 500 copies sold — before the product had even reached customers.
Mentorist: over 500K users from Facebook Ads
An app, not a course, but the same attention market and the same rule: whoever optimises every step of the journey wins. More than 500 A/B experiments made it possible to separate what works from what merely looks good.
What we do differently from a typical agency
The difference isn't in campaign settings — these days those are largely up to the algorithm. It's in the number of variants we can produce and test on the same budget, and in how quickly the lessons from those tests make it back to the sales page.
That shows up in the average results of our projects:
- average drop in CPA in the first 60 days
- -30%average drop in CPA in the first 60 days
- ROAS increase after rolling out AI creatives
- +150%ROAS increase after rolling out AI creatives
- more creatives tested on the same budget
- 10xmore creatives tested on the same budget
Common questions about Facebook ads for online courses
Let's run this on your numbers
Show us your course price, margin and what's happening in your campaigns today. We'll come back with the CPA that marks your break-even point and whether it's worth launching ads at your scale right now — even if the answer is "not yet".
Book a free consultation