The course analytics metrics worth tracking fall into two groups: learner metrics (enrollment, engagement, progress, completion) and business metrics (conversion, revenue, refunds, lifetime value). Track a handful from each group, define each one in writing, and attach every number to a decision you will actually make. A dashboard with 40 charts and no decisions is decoration.
This guide is for creators, coaches and educators who sell courses and other digital products. It explains what each metric tells you, how to calculate it, what can distort it and how to turn it into changes to your course.
TL;DR
- Measure two things at once: whether learners are getting value (engagement, progress, completion) and whether the product is healthy as a business (conversion, revenue, refunds, repeat purchases).
- Metrics are numbers. Analytics is the interpretation that leads to an action. Write down the decision each metric supports before you track it.
- Calculate completion only for learners who have had enough time, and leave out free enrollments, or the figure will mislead you.
- Treat refunds and chargebacks as guardrails next to revenue. A sales spike with rising refunds is not a win.
- Keep sales, payments, customers and learner activity in one place if you can. Reconciling several tools is where most tracking gaps come from.
Table of Contents
- Metrics versus analytics: a distinction that saves you time
- Three layers of course metrics
- Learner metrics: enrollment, engagement, progress, completion
- Did learners actually get the result?
- Business metrics: is the product attracting buyers and earning money?
- A starter dashboard: ten metrics, each tied to a decision
- How to read the numbers without fooling yourself
- Turning metrics into improvements
- What to look for in analytics tools
- Our perspective
- Run your course and your numbers in one workspace
- FAQ
- Sources
Metrics versus analytics: a distinction that saves you time
People use the two words as synonyms, but the difference changes how you work. In the academic literature, learning analytics is defined as the measurement, collection, analysis and reporting of data about learners and their contexts, in order to understand and optimize learning. Metrics are the specific numerical approximations of progress or achievement that tell you whether objectives are being met.
In practice: "41% of enrollees finished module 3" is a metric. Asking why the rest stopped, looking at what module 3 asks of them and rewriting its first lesson is analytics. The same source makes the point that the cycle is incomplete until the metrics drive an intervention that affects learners.
That is why this article keeps asking one question of every number: what would you change if it moved?
Three layers of course metrics
The same research groups learning metrics into three layers. They map neatly onto a creator's dashboard:
| Layer | What it covers | Creator examples |
|---|---|---|
| Efficiency (micro) | Individual learner activity | Lesson views, time on task, logins, last activity date |
| Effectiveness (meso) | Constructs such as engagement and progress | Percent of course completed, quiz attempts, replies in community |
| Outcome (macro) | Results at the end | Completion, certificate or project submission, testimonials, results learners report |
Most creators only look at the bottom row because it is easiest to collect. The top row is cheap but noisy: a learner who leaves a video playing is not engaged. The outcome row is the one your reputation depends on, and it is the one most often skipped. If you want a deeper grounding in how feedback and next steps help learners, study feedback tools for educators covers the learner side of the same idea.
Learner metrics: enrollment, engagement, progress, completion
Enrollment and activation
Enrollment counts the people who have access. It is not the same as sales if you offer free lessons, bonuses or comped seats, so keep paid and free enrollments separate from day one.
A more useful companion is activation: the share of new learners who complete a first meaningful action, such as finishing lesson one within seven days. Pick the action yourself, but write it down. If activation is low, your onboarding email and first lesson are the place to work, not your sales page.
Engagement
Engagement metrics include lesson starts, time spent, active days per week, comments and assignment submissions. Treat time spent with caution. A long session can mean deep work or a confusing lesson. Pair it with something that shows intent, like a submitted exercise.
Useful engagement questions:
- Which lesson has the highest drop-off compared with the one before it?
- Which lessons are replayed most? Replays often mean valuable or confusing.
- How many days pass between enrollment and first activity?
Progress
Progress is the percentage of lessons or modules completed per learner. Look at the distribution, not the average. If half your learners sit at 0 to 10% and the rest are above 80%, the average of 45% describes nobody. Segment by enrollment month so you compare learners who have had the same time.
Completion rate
Completion is the most quoted course metric and the easiest to get wrong. Metabase's creator analytics guide recommends using a maturity window and excluding free enrollments when you calculate it. Here is a worked example:
- 300 people enrolled in the last year. 60 were free-preview signups and 90 enrolled in the last three weeks.
- Your window is 60 days after enrollment, so the 90 recent learners are not yet eligible, and the 60 free signups are excluded.
- That leaves 150 eligible paid learners. If 48 finished, completion is 32%.
The naive figure, 48 out of 300, would have said 16% and sent you rewriting a course that is performing fine. Choose the window based on how long the course takes a motivated learner, and keep it fixed so month-to-month comparisons mean something.
Did learners actually get the result?
Completion tells you people reached the end, not that the course worked. Instructional design calls the deeper check confirmative evaluation: determining whether instruction achieves its long-term goals and solves the need it was designed for, beyond immediate success.
For a creator, that can be simple:
- State the result in one sentence before you build ("a learner can publish their first newsletter").
- Ask for evidence at the end: a submitted project, a before-and-after self-rating, a short survey.
- Follow up 30 to 90 days later with one question: did you use this, and what happened?
- Count the replies and read them. Ten specific answers beat a 4.8 star average.
Other industries face the same retention challenge. This retention playbook for dance studios shows how structured early check-ins reduce dropouts, an idea that transfers to the first weeks of a course.
Business metrics: is the product attracting buyers and earning money?
Learner metrics tell you whether the course is good. Business metrics tell you whether you can keep making it.
Conversion rate
Conversion is purchases divided by visitors at the step you care about: sales page to checkout, checkout to payment, lead magnet to purchase. Funnel tools commonly report initial conversion and step-by-step conversion with drop-offs, as in RevenueCat's funnel documentation. The step with the largest drop is where to start. If you are building the funnel itself, our guide to a course sales funnel, stage by stage maps each step.
Revenue, average ticket and recurring revenue
Track total revenue, average order value and, if you sell memberships or subscriptions, monthly recurring revenue (MRR). One research roundup lists MRR, lifetime value, acquisition cost, conversion and repeat purchase rates as the core KPIs for whether a digital product is attracting customers and producing sustainable profit.
Customer acquisition cost
CAC is total marketing and sales spend divided by new paying customers. Educational marketers sometimes call it student acquisition cost and also track cost per lead, which is recruitment spend divided by leads generated, as this education marketing guide explains. Compare CAC with revenue from a fixed window rather than treating conversion alone as success, as RevenueCat's paid campaign guidance advises.
CAC alone is also incomplete. The Profit CMO argues that profitability depends on collected revenue, delivery and sales costs, payment costs, support burden, refunds and expected future value. For a solo creator, a rough version works: revenue collected, minus ad spend, tool costs and payment fees, per sale.
Lifetime value and repeat purchases
Lifetime value is the average total revenue per customer over time. If you sell a $49 ebook, a $299 course and a $29 template pack, customers who buy twice are worth far more than the first order suggests. Repeat purchase rate tells you whether your follow-up emails and next offers are working. A tripwire offer is one way to create that second purchase early.
Refunds and chargebacks
Refund rate is the quality check on your revenue. Google Analytics documents refund events alongside purchases, and RevenueCat recommends watching refunds, chargebacks, cancellations and support contacts as guardrails next to campaign revenue. If a promotion lifts sales and refunds together, the promise on your sales page may be ahead of the course.
A starter dashboard: ten metrics, each tied to a decision
The guidance on avoiding "dashboard vanity" is to make sure every metric maps to a clear business decision, a defined value event and an owner who acts on it, according to this digital product analytics guide. For a solo creator, you are the owner of all of them. The table below is a sensible starting set.
| Metric | How to calculate it | Decision it supports |
|---|---|---|
| Sales page conversion | Purchases / sales page visitors | Rewrite the page, change the offer or traffic source |
| Checkout completion | Payments / checkouts started | Simplify checkout, check payment methods |
| Activation rate | New learners finishing lesson one in 7 days / new learners | Fix onboarding emails and lesson one |
| Lesson drop-off | Learners starting lesson N / starting lesson N-1 | Split, shorten or rework the weakest lesson |
| Mature completion rate | Finishers / paid learners past the window | Decide whether to change the course structure |
| Outcome evidence | Learners reporting the promised result / respondents | Update the promise, gather testimonials |
| Refund rate | Refunds / purchases in the same period | Adjust promise, add support, review the guarantee |
| CAC | Spend / new paying customers | Keep, cut or scale a channel |
| Repeat purchase rate | Customers with 2+ orders / customers | Build the next offer |
| MRR (if recurring) | Sum of monthly recurring charges | Plan capacity and pricing |
Review learner metrics weekly while a cohort runs and business metrics monthly. Looking at conversion daily mostly measures noise.
How to read the numbers without fooling yourself
Keep definitions fixed. Write down what counts as a customer, an active learner, a completion and a refund, the revenue basis (gross or net of fees) and the attribution window. RevenueCat's guidance recommends documenting exactly these and comparing equally mature cohorts.
Use consistent campaign tags. If one email link is tagged "newsletter" and another "Newsletter", they show up as two sources. Pick a naming convention for UTM values once.
Check the data. Missing events are common. Google's ecommerce setup recommends sending currency with value data and using transaction IDs on purchase and refund events, so refunds attach to the right sale. Reconcile orders, payments, access and refunds against your payment processor payouts now and then. If the numbers differ, trust the processor and find the gap.
Mind the context. A launch week, a holiday and a big email send each distort the month around them. Annotate your dashboard with what you did, otherwise you will not remember why February looked odd.
Beware small samples. With 30 learners, one person changes completion by more than three points. Use the numbers to form hypotheses and the learner conversations to test them.
To choose which numbers actually reflect value rather than activity, this piece on content effectiveness metrics offers a useful staged way of thinking.
Turning metrics into improvements
Analytics pays off only when it changes something. A practical loop:
- Pick one problem. The biggest drop in the funnel or the course, not five.
- Form a hypothesis. "Lesson 4 asks learners to set up a tool before they see why."
- Change one thing. Move the payoff earlier, add a worksheet, shorten the video.
- Compare equal cohorts. Same maturity window, same traffic source where you can.
- Write down the result and the next question.
Some examples of metric patterns and the fixes they point to:
- High sales page traffic, low conversion: the offer or the promise is unclear. Revisit who the course is for and what they will be able to do.
- Good conversion, weak activation: buyers are not getting started. Send a welcome email with one first step and see whether activation moves. Our guide to course lead nurturing covers the sequence.
- Strong activation, steep drop at one lesson: rework that lesson before touching anything else.
- High completion, few repeat purchases: the course works but there is no next step. Design the follow-up offer and the email that introduces it.
- Rising refunds: compare what the sales page promises with what lesson one delivers, and add a clearer start path.
- Steady sales only during launches: consider an evergreen funnel so sales do not depend on a single send.
What to look for in analytics tools
You do not need an enterprise stack, but you do want a few capabilities. According to one guide to funnel tools for digital products, look for clear funnel step logic, identity stitching so one person is recognized across sessions, raw data export and event-based tracking rather than page views alone.
A short checklist:
- Can it show paid sales, refunds and learner activity for the same person?
- Can you export the raw data if you outgrow it?
- Can it track events (lesson completed, checkout started), not just page visits?
- Does it let you segment by source and by enrollment date?
- Will you actually open it every week? A simple view you use beats a powerful one you avoid.
The more tools your sales, email and course content sit in, the more reconciling you do. Many creators end up with learner data in one place, payments in another and email lists in a third, and the numbers never quite match. Keeping them together is the simplest way to avoid that.
Our perspective
We built Genesize around a simple observation: creators lose track of their numbers because their work is scattered. The course lives in one tool, payments in another, the email list in a third, and nobody can answer "which email led to which sale, and did that buyer finish the course?"
That is why we put product creation, Stripe payments, CRM, email follow-ups and analytics in a single workspace. When a lesson, a buyer and a follow-up email are connected, the metrics in this article stop being three exports you stitch together by hand and become one picture you can act on.
Our view is that analytics should be small and decision-led. Know your conversion, your activation, your mature completion and your refunds. Then use what you learn to shape the next lesson, the next email, the next offer. Our platform is designed to give you AI-powered growth recommendations alongside your data, so the step from "what happened" to "what should I change" is shorter.
Start with five metrics, review them on a schedule and change one thing at a time. That habit matters more than any tool.
Andrew
Run your course and your numbers in one workspace
Genesize is a platform for creating AI-assisted digital products such as ebooks, courses and resource packs, publishing them to simple storefronts, accepting payments through Stripe, managing customers and automating email follow-ups, with analytics in the same workspace. Plans are Starter at $24 per month, Creator at $44 per month and Business at $84 per month, and each includes a free trial first.
You can see the plans and start on the Genesize pricing page. If you are planning your first release, the 6-week course launch plan shows which numbers to watch at each stage.
FAQ
What is a good course completion rate?
There is no universal benchmark, because it depends on price, course length, audience and how you define completion. Compare yourself with your own past cohorts, using the same maturity window and excluding free enrollments. Improvement from one cohort to the next is more informative than any outside figure.
Which metrics matter most for a first-time course creator?
Start with sales page conversion, activation (first lesson finished), lesson-by-lesson drop-off, mature completion and refund rate. Those five show whether people buy, start, continue and are satisfied. Add CAC and repeat purchase rate once you run paid traffic or sell more than one product.
What is the difference between course analytics and digital product analytics?
Course analytics focuses on learners: engagement, progress and outcomes. Digital product analytics adds the commercial side, such as funnel conversion, revenue, refunds and customer value, and covers products beyond courses like ebooks and templates. Creators need both views, ideally connected to the same customer.
How often should I review my course analytics?
While a cohort is running, check learner metrics weekly so you can help people who stall. Review business metrics monthly, or after any launch or promotion. Daily checks tend to produce reactions to noise rather than decisions.
Do I need Google Analytics to track a course?
Not necessarily. Google Analytics can track ecommerce events such as purchases and refunds, but a platform that records sales and learner activity together may give you what you need without extra setup. If you do use it, send currency with value data and transaction IDs with purchases and refunds.
Sources
- Learning Analytics, Springer Nature Link
- Instructional Design Evaluation, Social Sci LibreTexts
- Creator Analytics in Metabase
- Analyzing Funnels, RevenueCat
- How To Measure Paid Web-To-App Campaigns, RevenueCat
- Measure ecommerce, Google for Developers
- Course Marketing Analytics and Performance, The Profit CMO
- Educational Marketing Metrics, mkt4edu
- EdTech Metrics That Matter
- Digital product analytics: the practical version
- Funnel analytics tools for digital products