Meta ads to sell online courses: Instagram and Facebook
When Meta ads make sense for a video course, how to set a maximum cost per sale, why to optimise for sign-ups, not purchases, and how the learning phase works.
Sooner or later every creator wonders whether it's time to pay to reach more people. Instagram and Facebook ads can bring new students, but they can also burn through a launch's revenue in a few days.
This guide is for creators who sell video courses and want to know whether, how and how much to invest in Meta ads. It doesn't walk through Ads Manager menus, which change often. It covers the maths to do first and the system rules that don't change.
When does paying for ads make sense for a course?
When three things are true: the course has already sold to your organic audience, you know your margin per sale and you have a path that turns a stranger into a buyer (free content, emails, a launch). If one is missing, ads usually cost more than they return.
The reason is simple: ads bring people who don't know you. In other words, they start the relationship rather than finish it. Someone who has followed you for months buys after one email; someone seeing you for the first time in an ad needs time and proof. Ads work best as the first step of a journey, not a shortcut to the sale.
What's the most you can pay for a sale?
Your margin per sale, minus the profit you want to keep. Margin is price minus payment fees and other per-sale costs. With our course profit calculator: a $149 course with 5% plus 30 cents in fees leaves $141.25 per sale. If you spend $141.25 in ads per sale, you break even on that sale but haven't covered production costs yet.
See the data as a table
| Ad cost per sale | Margin after ads |
|---|---|
| $20 per sale | 121.25 $ |
| $50 per sale | 91.25 $ |
| $80 per sale | 61.25 $ |
| $120 per sale | 21.25 $ |
Therefore, set your maximum cost per sale before you start and write it down. Then the comparison during the campaign is simple: if the real cost stays above it for more than a few days after the learning phase, stop or change something. To price the course from margin, see how much to charge for an online course.
What is the learning phase and why does it change your budget?
It's the period when Meta's delivery system explores the best way to deliver a new ad set or one with a significant edit. According to Meta's Business Help Centre, ad sets exit learning once performance stabilises, usually after about 50 results in the week after the last significant edit. If an ad set can't get enough results, Meta labels it "Learning limited".
In particular, during learning, Meta says, ad sets are less stable and usually have a higher cost per action. Editing ads, budgets or audiences in this phase resets learning.
- Results, roughly, to exit the learning phase
- 50
- Days: the window after the last significant edit
- 7
- Dollars a week for 50 purchases at $40 each (example)
- 2,000
Optimise for purchases or for sign-ups?
This is the most important choice for a creator. If you ask Meta to optimise for course purchases and each costs $40, you need about $2,000 a week for the ad set to collect the 50 results Meta mentions. However, few creators have that budget for a single course.
If you optimise for sign-ups to free content instead, and each sign-up costs a few dollars, 50 results a week is reachable on a much smaller budget. Those sign-ups join your list and the sale happens later, through launch emails. That's why many creators use ads to grow their list rather than to sell directly. How to build the free content is in lead magnet ideas for creators.
An example with numbers
Take a $149 course and a budget of $300 a week. If you optimise for purchases and each purchase costs $40, the ad set collects about 7 results a week: far from the 50 Meta mentions, so it stays in limited learning and costs stay unstable. If instead you optimise for sign-ups to a free lesson and each sign-up costs $3, the same $300 brings about 100 sign-ups a week, enough to exit the learning phase. In that case, however, the sale doesn't happen in the ad: it happens later, in your emails. As a result, the number to watch isn't cost per sign-up on its own, but how many of those people buy at launch. For example, if 3% buy, 100 sign-ups become 3 sales, or $100 of ads per sale: below the $141.25 margin. The costs are assumptions: replace them with yours.
How does the ad auction work?
Every time there's a chance to show someone an ad, Meta runs an auction among the ads whose audience includes that person. According to Meta's Business Help Centre, the winner is the ad with the highest total value, combining three factors: the advertiser's bid, the estimated action rates (the probability that the person takes the desired action) and ad quality, measured partly through feedback from people who see or hide the ad.
The practical consequence, in Meta's own words: a more relevant ad can win against ads with higher bids, and clickbait or engagement bait don't improve performance. For creators, therefore, content that already works organically often makes the best ads.
Which ads work for a course?
Ads that look like your best content, not like advertising. A 30-60 second video where you teach something useful, ending with an invitation to the free resource, is a natural format for course creators. Start from content that has already performed on your profile, which you can spot with the right numbers: see creator metrics that matter.
Three practical rules from Meta's help page on learning: don't create too many ads and ad sets (the system learns less about each), use realistic budgets and avoid editing until learning is over.
How do you build a simple campaign?
A minimal structure for a creator has two parts:
- Ads to free content, aimed at new audiences, optimised for sign-ups. They grow your list steadily.
- Launch ads, aimed at people who've already engaged with you (site visitors, video viewers), pointing to the sales page or webinar. Small audiences, but already warm.
In addition, connect everything to a path: free content, welcome emails, launch. Without the path, people from ads sign up and forget you. The full path is in how to launch an online course.
How do you measure whether ads pay off?
Measure cost per sign-up, the share of sign-ups who buy at launch and the resulting cost per sale. Example: if a sign-up costs $3 and 2% of sign-ups buy, each sale costs about $150 in ads. With a $141.25 margin, you're losing money. You have three levers: lower cost per sign-up, raise the share who buy (with better emails) or raise each customer's value.
The 2% is an assumption for the example, not an average: measure your real figure, which you'll only know after a first launch.
Which mistakes should you avoid?
- Advertising a course that has never sold. Ads don't fix an offer that doesn't convince.
- Optimising for purchases on a small budget. The ad set stays in limited learning.
- Changing everything daily. Each significant edit restarts learning.
- Only watching cost per click. Cost per sale is what counts.
- Overblown promises. Besides breaking ad policies, they bring refunds.
Key terms. The learning phase is when Meta explores how to deliver an ad set. Cost per result is your average spend for each optimised action (a sign-up, a purchase). Total value is the score Meta uses to pick the auction winner.
How we checked. We read Meta Business Help Centre pages on the learning phase and on ad auctions. Costs per sale and per sign-up are our own examples, not market averages: we found no independent data on ad costs for video courses. Contact: hi@heytea.app. Editorial note: we check every figure against the sources listed below and correct mistakes openly.
Frequently asked questions
Is it worth running Instagram ads for an online course?
It's worth it once you know the course sells to your organic audience and you know your margin per sale. Ads amplify something that works; they don't fix an offer that doesn't convince. Before spending a cent, work out the most you can pay for a sale without losing money.
What budget do course ads need?
It depends on the event you ask Meta to optimise for. Meta says an ad set usually exits the learning phase after about 50 results in the week after its last significant edit. If the result is a purchase and each purchase costs $40, that's about $2,000 a week, which is why many creators optimise for email sign-ups instead.
Should ads go to the sales page or to free content?
For people who don't know you, usually to free content: a lead magnet, a webinar, a lesson. People who've never seen you rarely buy a course on first contact. Those who sign up join your list and get your launch emails, which is where the sale happens.
What is the Meta ads learning phase?
The period when Meta's delivery system explores how best to deliver a new ad set or one with a significant edit. During it, performance is less stable and cost per result is usually higher. Meta advises against editing ads until the learning phase is over.
How does an ad win Meta's auction?
Not simply by bidding highest. Meta explains that the winning ad has the highest total value, which combines the bid, the estimated action rate (how likely the person is to take the desired action) and ad quality. A more relevant ad can beat higher bids, while clickbait and engagement bait don't improve performance.
Sources
- Meta Business Help Centre, "About the learning phase". Accessed 18 October 2026. facebook.com
- Meta Business Help Centre, "About ad auctions". Accessed 18 October 2026. facebook.com
- Tea free tool, course profit calculator: formulas for margin per sale. heytea.app