Almost every course creator tracks the same three things: total sales, total revenue, and follower count. Those numbers feel good in a launch week, but they tell you nothing about why a course sold, why a second launch did worse, or why your refund requests doubled after you raised your price by $50. Revenue is an outcome. Metrics are the causes. If you only watch revenue, you are reading the scoreboard after the game has ended.

Here is the advantage of running a course business on a platform you control: you get to see the whole path a buyer walks, from the first anonymous visit to the last completed lesson. Marketplace sellers receive an aggregated payout report and call it data. When you own the storefront, you own the funnel — every visit, every abandoned checkout, every student who stops watching in module three — and that is what lets you fix things instead of guessing.

Below are 15 metrics, grouped into four parts that follow the same sequence your students follow: attention, conversion, delivery, and compounding. For each one you get a plain definition, why it moves revenue, how to calculate it, a healthy range, and the specific action to take when the number is off. At the end you will find a one-page weekly dashboard and a diagnostic order so you know which metric to fix first.

Why your numbers look better (and mean more) when you own the storefront

Marketplace analytics are built for the marketplace, not for you. You usually get sales count, a rough geographic breakdown, and a star rating. You rarely get visitor counts for your own sales page, you cannot see where the checkout broke, you cannot email your buyers without asking permission, and you cannot A/B test a price. The moment your course lives on your own domain, all of that changes.

Owning the storefront also changes the economics of the numbers you can measure. On your own site, revenue per visitor is a decision you can act on, because you control the sales page, the checkout flow, the bumps and bundles, and the post-purchase sequence. Attention is a compounding asset: an email subscriber you capture on your own site can be marketed to for years, while a follower on a rented platform can be deprioritized by an algorithm tomorrow.

The catch is that a self-hosted course business gives you no default dashboard. Stripe shows payments, your email tool shows subscribers, your course platform shows lesson progress, and your analytics tool shows sessions. Nobody joins those four datasets for you. That is the work — and it is the reason most creators who move off a marketplace feel less informed before they feel more informed.

Set up these four data sources before you need them

Every metric below can be calculated from four sources you probably already have. Wire them together once, and the numbers refresh themselves.

• Web analytics — sessions, unique visitors, traffic sources, referral paths, and scroll or engagement events on each page. A free analytics property plus consistent UTM tagging covers the attention metrics.
• Payment processor reports — gross volume, transaction count, average order value, refunds, and disputes. This is your financial source of truth; platform dashboards sometimes net out fees and hide the real average order value.
• Email platform — list size, net growth, lead magnet conversion, and deliverability. Your list is the only audience asset that cannot be taken away by an algorithm change.
• Course platform — enrollments, lesson-level completion, time to first lesson, and support requests. If your platform cannot report lesson-level completion, you cannot measure the delivery metrics in Part 3, and you should treat that gap as a product requirement.

A practical setup note: agree on one definition for each metric and write it down. “Conversion rate” that sometimes includes free trials and sometimes does not will send you chasing ghosts every month. One number, one definition, one cadence — weekly at most, and only after you have enough traffic for the ratio to mean anything.

One more rule before the list: never draw conclusions from fewer than about 100 conversions or 1,000 visitors in the measured period. Below those thresholds you are reading noise. Track the metric, but do not redesign your business around a 2.1% conversion rate calculated from 19 sales.

Part 1: Attention — are the right people arriving?

Attention metrics answer a narrow question: how many people who could plausibly buy your course actually reach the page where buying happens? They are not vanity metrics — they are the denominator of every ratio you will calculate later. Fix attention and conversion together; fixing one alone rarely moves revenue.

1. Sales page unique visitors

What it is: the number of distinct people who loaded your course sales page in a period, not the number of page views. Why it predicts revenue: it is the top of your funnel, and it is the one lever with no ceiling — a 30% lift in qualified visitors produces a 30% lift in sales at constant conversion. How to track it: a unique page-view or user count filtered to the sales page URL.

Healthy range: any number that grows month over month is healthy; the ratio that matters is qualified visitors. A thousand visitors from a Facebook group of hobbyists may be worth less than a hundred from a niche newsletter. If it is flat: you have a distribution problem, not a conversion problem. Publish consistently, build one evergreen lead magnet, and get on podcasts or in communities where your buyers already gather.

2. Traffic source mix (owned versus borrowed)

What it is: the percentage of your sales page visitors that arrive from channels you control — email, search, direct, and referrals from partners — versus channels you rent. Why it predicts revenue: borrowed attention is volatile. A single algorithm change on a social platform routinely cuts reach by half, and creators who depended on it lose a third of their revenue overnight. Owned traffic is what makes revenue forecasting possible.

How to track it: segment traffic by source in your analytics, then label each source owned or borrowed and calculate the split. Healthy range: aim for 50% or more of pre-sale traffic from owned or earned sources. If it is skewed to borrowed: add a lead magnet and a low-commitment email sequence that converts social visitors into subscribers, so the same attention pays you twice — once in the post, again in the launch. If your courses currently live on a rented platform, it is worth reading the proven steps for moving a course onto your own site before you build another launch on a channel you do not control.

3. Net email list growth rate

What it is: new subscribers minus unsubscribes and bounces, expressed as a percentage of the list per period. Why it predicts revenue: for most course businesses each active subscriber is worth somewhere between a dollar and several dollars a month in eventual revenue, so list growth is a leading indicator of launches three and four. How to track it: (new − lost) ÷ starting list size, measured weekly.

Healthy range: 1% to 3% net monthly growth for an established list, much higher in the first year when the base is small. Watch for gross adds masking a rising churn rate — a list that adds 500 and loses 480 is not growing, it is churning. If growth is negative: audit your welcome sequence first, since most unsubscribes happen in the first three emails, then reduce send frequency or tighten your lead magnet’s promise so it attracts people who actually want the course.

4. Lead magnet or waitlist conversion rate

What it is: the share of visitors to your opt-in or waitlist page who hand over an email address. Why it predicts revenue: it measures whether your audience understands the transformation you sell. If 2% of visitors want your free resource, a paid course four times the commitment is unlikely to convert many of them. How to track it: form submissions ÷ unique visitors on the opt-in page.

Healthy range: 20% to 40% for a dedicated landing page sending one coherent promise, with ads or cold traffic trending lower. If it is under 10%: the problem is usually specificity, not design. Swap “free email course” for the actual outcome (“the 12-lesson path from blank page to a finished module plan”) and test a single-field form.

Part 2: Conversion — where money is won or lost

Conversion metrics tell you how efficiently attention turns into cash. Nine out of ten underperforming course businesses have a conversion problem rather than a traffic problem, and the fix is usually cheaper and faster because you already paid for the visitors.

5. Sales page conversion rate

What it is: purchases divided by unique visitors to the sales page, in a defined window. Why it predicts revenue: it is the multiplier between attention and income, and small changes compound. Moving from 1% to 2% conversion on the same traffic doubles revenue with zero extra spend. How to track it: completed purchases ÷ unique sales page visitors, both counted for the same date range.

Healthy range: 1% to 3% for warm traffic you own; 3% to 10% is achievable for a focused launch to a list that has been educated over weeks. Cold paid traffic usually lands under 1%. If it is under 1%: check the offer before the page. Ask whether the outcome is specific, whether the price matches perceived transformation, and whether the page answers the five objections every buyer has: time, money, ability, support, and results. Two structural wins worth testing: a curriculum breakdown with module outcomes, and a short refund policy instead of a paragraph of hedging. The elements of a high-converting course landing page are worth auditing against your current page.

6. Checkout completion rate

What it is: completed purchases divided by checkout sessions started. The inverse is cart or checkout abandonment. Why it predicts revenue: these are buyers who already decided yes. Losing them is the cheapest revenue you will ever recover. How to track it: checkout-initiated events from analytics against successful transactions from your payment processor.

Healthy range: 65% to 85% completion. If it is under 60%: look at friction, not persuasion. Unexpected costs at the last step, forced account creation, a payment method you do not accept in your buyers’ country, or a slow video-heavy page that times out on mobile. Offer a local payment method if your audience’s most common cards are declined, and always test the purchase flow yourself on a phone with a real card.

7. Average order value

What it is: total revenue divided by the number of transactions — including bundles, payment plans, and any order bump. Why it predicts revenue: it raises revenue per visitor without needing more traffic or a higher headline price, and it is much easier to move than conversion. How to track it: gross revenue ÷ transaction count from your payment processor, not your platform’s net report.

Healthy range: there is no universal number — the target is upward movement over quarters. If it is flat: add a complementary low-priced product at the point of purchase (templates, a worksheet pack, a live Q&A), offer a bundle of two related courses at a modest discount, or introduce a three-payment plan for courses above roughly $300. Payment plans usually lift conversion more than average order value, and both lift total revenue. If you are about to raise the headline price instead, read up on how course pricing psychology shapes perceived value first.

8. Revenue per visitor

What it is: total revenue divided by total unique visitors across all pages in a period. It is the single most decision-useful number in a course business because it folds conversion, price, and average order value into one figure. Why it predicts revenue: it tells you exactly what one more visitor is worth, which is the number you need before you spend a dollar on advertising. If revenue per visitor is $1.20, paying $1.50 per click is a slow leak; if it is $4, paying $1.50 is a machine. How to track it: gross revenue ÷ unique visitors, weekly or monthly.

Healthy range: $0.50 to $2.00 for broad content sites, $2.00 and up for a well-segmented list. If it is falling while revenue holds: you are buying growth with volume rather than efficiency. Fix the weakest of its three inputs — conversion, price, order value — rather than adding traffic.

9. Refund and chargeback rate

What it is: refunds divided by transactions in the same period, plus disputes separately. Why it predicts revenue: refunds are a lagging signal of mismatched expectations, and they cost more than lost revenue — processing fees are rarely returned and a high dispute rate jeopardizes your ability to accept cards at all. A 10% refund rate quietly erases a price increase. How to track it: refund volume ÷ transaction count, monthly, plus a log with a one-line reason code for each refund.

Healthy range: under 5% for most digital courses; under 3% is excellent. If it is climbing: read the reason codes before changing anything. Refunds clustered on lesson three are a delivery problem, refunds clustered in the first 48 hours are a positioning problem, and refunds clustered after a pricing change are a value-perception problem. Fixing the matching cause is worth more than any refund policy ever will be.

Part 3: Delivery — does the course do what the sales page promised?

These three metrics are the most neglected in the entire industry, because they do not appear on any payment report. They are also the strongest predictor of whether you get to keep selling: course businesses do not usually die from bad traffic, they die from a first cohort that quietly does nothing, refunds, and never buys again.

10. Activation rate

What it is: the share of new students who complete the first lesson or module within seven days of purchase. Why it predicts revenue: activation is the earliest reliable forecast of completion, satisfaction, refunds, testimonials, and repeat purchases. Students who never start almost never finish, and they almost never leave the review you need for the next launch. How to track it: count recent buyers, then count how many logged a completed first lesson within a week. Your course platform’s student and progress views make this a two-minute query.

Healthy range: 60% to 80%. If it is under 50%: the gap is almost always in the first ten minutes. Send an immediate welcome email with a single “start here” link, keep the dashboard uncluttered, and make the first lesson short, specific, and immediately useful — no history of the topic, no long introductions. Ask for the first small win before asking for a big time commitment.

11. Drop-off lesson

What it is: the lesson with the largest fall in completion — the point where the cohort leaks. Why it predicts revenue: a single lesson can be responsible for a 20-point swing in completion, and completion correlates with refunds, referrals, and upgrades. In a multi-module course, one badly paced lesson can cost more revenue than a weak email subject line ever will. How to track it: plot lesson completions in order and find the steepest step down; then look at time on lesson and support questions about that specific topic.

Healthy range: no single step losing more than 15% of the students who reached the previous lesson. If one lesson leaks badly: suspect length before content. Split a 25-minute lesson into two, add a summary in the first 30 seconds, or replace a theory-heavy lesson with a worked example. Video length and lesson order are the two cheapest levers in course design.

12. Completion rate and support load

What it is: the percentage of students who finish the course, paired with the number of support requests per hundred students. Why it predicts revenue: completion rate drives outcomes, and outcomes drive testimonials, referrals, and the launch you run next quarter. Support load per hundred students is the honest measure of whether your course is self-explanatory: a rising ratio means new students keep hitting the same wall. How to track it: completions ÷ enrollments for a fixed cohort, and support tickets ÷ students, tracked monthly.

Healthy range: 25% to 50% for self-paced courses; above 50% is excellent. If completion is low: two structural fixes outperform nudging. First, build deadlines or cohort pacing into the course so students move on a schedule instead of “whenever”. Second, add a mid-course checkpoint — a short submission, quiz, or live call — because the moment a student has to produce something, they stop drifting. For support load, the fix is documentation rather than more answering: when three people ask the same question, that answer belongs inside the course, not in an inbox. If completion is your weakest number, the proven student retention strategies work through the same problem from the delivery side.

Part 4: Compounding — do students come back and bring others?

The last three metrics determine whether your business gets easier or harder each year. A course business that acquires new students forever, with no reuse of the ones it already has, is a treadmill. These numbers are the engine.

13. Repeat purchase and attach rate

What it is: the share of existing students who buy a second product within 12 months, plus the average number of products per customer. Why it predicts revenue: existing students convert several times better than cold traffic, cost nothing in ad spend to reach, and are already educated on your teaching style. A course business with two products and a 20% attach rate grows faster than one with five products and no follow-up. How to track it: count customers with two or more transactions ÷ total customers, rolling 12 months.

Healthy range: 15% to 30% attach within a year for creators with an obvious next step. If it is near zero: the fault is usually sequencing. Offer the next logical outcome at the moment of the first win, shortly after a student completes the course, and give the second product a clear “what comes next” relationship to the first. Email the completion milestone — it is the single highest-intent moment you will ever have with that student.

14. Referral and word-of-mouth share of sales

What it is: the percentage of new enrollments that came from an existing student’s recommendation, an affiliate link, or direct sharing. Why it predicts revenue: referred buyers convert at higher rates, refund less, and lower your blended acquisition cost permanently. A course with 25% referral traffic can be scaled profitably with paid ads; one with 2% is fragile. How to track it: a “how did you hear about this?” field at checkout plus affiliate or partner link tracking.

Healthy range: 15% to 30% of enrollments within a year of launch. If it is low: ask for referrals at the right moment — right after a stated win or a completed final lesson — and make sharing transactional: a partner program with a straightforward commission, a template your students will send to colleagues, or a certificate they are happy to post. Referrals follow results, so check metrics 10 through 12 before investing here.

15. Lifetime value to acquisition cost ratio

What it is: total revenue from a customer over their life divided by the cost of acquiring them. If you spend nothing on acquisition, substitute the value of the time and content it takes to win a student. Why it predicts revenue: this ratio decides how aggressively you can grow. At 1:1 you are working for every dollar; at 3:1 you can invest in ads, affiliates, and partnerships with confidence. It is also the number that justifies raising prices instead of always discounting. How to track it: (average order value × average purchases per customer) ÷ cost per acquired customer, reviewed quarterly.

Healthy range: 3:1 or better. If it is under 2:1: raise the numerator before cutting the denominator. Increasing attach rate and average order value lifts lifetime value without touching ad budgets, and it is usually faster than optimizing a campaign for a marginally cheaper click.

Your one-page weekly dashboard

Do not monitor all fifteen weekly. Four are weekly operating numbers; the rest are monthly or quarterly reviews. That split keeps you focused without drowning in spreadsheets.

• Weekly: sales page visitors, sales page conversion rate, revenue per visitor, new students activated in the first seven days.
• Monthly: traffic source mix, net list growth, checkout completion, refund rate, drop-off lesson, support load.
• Quarterly: attach rate, referral share, lifetime value to acquisition cost ratio.

A simple target sheet to keep next to the dashboard:

Five numbers you can safely ignore

There is a cost to watching the wrong things: you spend your attention on numbers you cannot influence. These five feel important and rarely are.

1. Follower count. It measures an audience you rent, not buyers you have. A 40,000-follower account with a 300-person email list is a smaller business than it looks.
2. Total page views. Without knowing which page and which source, aggregate views hide both your best traffic and your worst.
3. Average time on site. A long visit can mean engagement or a confused reader re-reading the pricing section. Use scroll depth and checkout behaviour instead.
4. Social engagement rate. Likes and shares are a distribution signal, not a demand signal. Your conversion rate and revenue per visitor already tell you if the audience matches the offer.
5. Last launch’s revenue in isolation. A single launch number with no denominator, no refund rate, and no completion data is a story, not a metric. Compare revenue per visitor across launches instead.

The order to fix things in

When several numbers are weak, the temptation is to improve everything at once. Work through this sequence instead. Each step changes the value of the ones below it.

1. Checkout completion first. If buyers cannot pay easily, every other improvement is capped. Fix payment friction, then move on.
2. Activation second. Students who never start will not finish, will not refer, and will not return. A welcome sequence and a short first lesson are fast wins.
3. Sales page conversion third. With payment and delivery sound, improve the offer and the page: specific outcomes, curriculum clarity, objection handling, proof.
4. Average order value fourth. Bundles, order bumps, and payment plans raise revenue from traffic you already have.
5. Traffic last. Only buy or chase more attention once revenue per visitor and lifetime value to acquisition cost justify it. Otherwise you are paying to widen a leaky funnel.

Frequently asked questions

How many visitors do I need before conversion numbers mean anything?

Roughly 1,000 unique visitors to a page and at least 100 conversions in the measured window. Below that, treat figures as directional only and avoid redesigning the page or changing the price based on a week of thin data.

What is a realistic first-year completion rate for a self-paced course?

Twenty-five to fifty percent is the normal band, and higher rates usually come from added structure rather than better video production. Cohort pacing, deadlines, and a mid-course checkpoint move completion more than re-recording lessons.

Which single metric should a new creator watch most closely?

Revenue per visitor. It combines traffic quality, page conversion, price, and order value into one number, and it tells you immediately whether a paid acquisition channel is viable or a slow loss. See the full cost breakdown of running a course storefront for the expense side of that equation.

Putting it all together

Fifteen metrics sounds like a lot until you sort them by cadence, and then it is four weekly numbers, six monthly ones, and three quarterly reviews. The sequence is always the same: verify that buyers can pay, verify that students actually start and finish, sharpen the offer and the page, raise order value, and only then buy more attention. Do that for two quarters and you will know your business better than the vast majority of creators — because you will be reading causes instead of celebrating outcomes.

A useful next step is to make sure the delivery half of your dashboard is even visible to you: you cannot improve activation, drop-off, or completion with a platform that only tells you how much money came in. Owwlish was built for exactly that gap — it puts your courses on your own website, keeps your payment flow simple, and shows you student-level enrollment and progress data so the metrics in Parts 3 and 4 are things you can actually measure and act on, instead of assumptions you hope are true. If you have been selling through a marketplace and only ever seeing a payout summary, start with a free Owwlish account and get the full picture of your own course business.