Instructional design

Instructional design is the work you do before you press record. You decide what a learner should be able to do at the end, what they need to learn to get there, in which order, and how they will practice. The videos, worksheets and emails come after, as the result of those decisions.

It concerns every creator, coach or expert who sells a course. Knowing a subject well does not mean you can teach it. Experts skip steps they no longer notice and pile on details a beginner cannot use yet. Instructional design is the discipline that closes that gap, so your knowledge turns into a skill in someone else's hands.

What is instructional design?

Instructional design is the structured process of creating learning experiences that lead to a measurable change in what people can do. It covers four questions. Who is learning and where do they start? What should they be able to do at the end? What sequence of content and practice gets them there? How do you and they know it worked?

It is not video production. Lighting, editing and slides are packaging. A beautifully shot course with no clear objective still fails. It is not content writing either. A blog post informs. A course has to change behavior, which requires practice, feedback and repetition. And it is not a corporate-only discipline. Big companies hire instructional designers, but a solo creator applies the same logic on a smaller scale.

Related vocabulary: learning objectives (what the learner will be able to do, written with an action verb), ADDIE (analyze, design, develop, implement, evaluate, a classic model), backward design (start from the final result and plan backwards), scaffolding (support that is removed as the learner progresses), formative assessment (small checks along the way). Each course is usually split into course modules, each module into lessons.

Why it matters

Poor design shows up in three places: students who stop halfway, refunds, and no word of mouth. People recommend courses that worked for them. They do not recommend courses they abandoned in week two.

A worked example. A fitness coach sells a $197 program to 200 students a year, so $39,400. The first version is 14 hours of video recorded in one long flow. About 15% of buyers finish, refunds reach 8% ($3,152 lost) and almost nobody posts results. She redesigns it: one objective per week, lessons of 10 minutes, a practice task at the end of each, a checklist per module. The content shrinks to 6 hours. Completion rises to 45%. Refunds fall to 3%, which saves about $2,000 a year. And the 90 students who finish share before-and-after photos, which bring 40 extra sales the next year, or $7,880. The redesign took two weeks.

Good design also protects your time. Clear lessons generate fewer support emails, and a well-structured course can be updated one lesson at a time instead of re-recorded from scratch.

How it works

Most creators do well with a simple backward process. Each step produces a short written output.

• Define the learner. Write the starting level, goal and constraints of your ideal student. One page is enough.
• Write the final objective. One sentence with an action verb: "Shoot and edit 20 catalog photos with a phone in one afternoon." Avoid "understand" or "know about", which you cannot observe.
• Break it into sub-skills. List everything the learner must be able to do to reach the objective. Usually 4 to 8 items. Each becomes a module.
• Order them. Start with what gives a first visible result quickly, then build on it. Prerequisites come before the skills that depend on them.
• Plan practice before content. For each module, decide the exercise first. Then add only the explanation needed to do that exercise.
• Choose formats. Video for demonstrations, text for reference, templates and checklists for doing, live sessions or email for feedback. Short units work well for busy learners, see microlearning.
• Build a pilot. Run the course with 5 to 20 people, watch where they get stuck, and fix those lessons first.
• Evaluate. Track completion, results reported by students, refund reasons and the questions people ask. Each recurring question points to a lesson to improve.

Benchmarks and examples

These ranges come from typical creator courses, not academic programs.

• A focused course that teaches one skill often runs 2 to 5 hours of video, split into 20 to 40 lessons of 5 to 12 minutes.
• Self-paced courses without design work often see 5% to 15% completion. With clear objectives, short lessons and practice tasks, 30% to 60% is achievable. See completion rate.
• A good rule is roughly one practice task for every 10 to 15 minutes of content.
• A language teacher selling a 89 euro pronunciation course cut 40 lessons down to 18 after a pilot showed that students skipped the theory. Sales stayed flat, reviews improved and the support inbox went from 30 emails a week to 8.
• A software trainer turned a 6-hour screen recording into 6 projects, each producing a working file. Students now finish because each project ends with something they can show.
• A small candle brand built a 25-minute customer course on safe burning and care. It is not sold. It reduces returns and questions, which is also instructional design.

Common mistakes

• Starting with content instead of objectives. Recording everything you know produces a long course with no clear finish line.
• Confusing length with value. Buyers pay for a result, not for hours. A 20-hour course often signals poor editing, not more value.
• No practice. Watching someone do a task is not learning to do it. Without exercises, students feel they understood and then fail alone.
• The curse of knowledge. Skipping steps that seem obvious to you. A pilot with real beginners reveals them in one session.
• Never revising. The first version is a draft. The courses that sell for years are rewritten after each cohort.

Best practices

• Write objectives with action verbs. "Build", "write", "set up", "choose". If you cannot observe it, rewrite it.
• Get a quick win in the first 30 minutes. An early visible result keeps people going through the harder middle.
• One lesson, one idea. If a lesson covers two ideas, split it. Short lessons are easier to finish, find and update.
• Give templates and checklists. A worksheet a student fills in is worth more than ten minutes of explanation.
• Add feedback loops. Office hours, a reply-to-this-email assignment or peer review. Feedback is what turns practice into progress.
• Show the path. A one-page course map with modules and outcomes lets people see where they are and what comes next.
• Improve from questions. Keep a list of every question students ask. Each one is a missing or unclear lesson.

In Roctify

Roctify is where you sell and deliver the course, not where you design it. You add your course or downloadable files as a digital product in the shared catalog, sell it from your link-in-bio page or your online store, and buyers get access automatically after payment through Stripe or PayPal, with 0% transaction fees on every plan. On the Creator plan and up, email marketing lets you send a short series of emails as campaigns to support the design, for example a welcome email with the course map and a weekly nudge with the next practice task.

Roctify does not include quizzes, certificates or student progress tracking, so assessment and feedback happen through assignments sent by email, forms or live calls you host elsewhere. Forms (Creator plan and up) are a simple way to collect a pilot group's feedback after each module.

FAQ

Do I need a degree to do instructional design?

No. Formal training helps in corporate settings, but creators can apply the core method on their own: define the learner, write the objective, plan practice first, test with a pilot. Most gains come from those four steps.

How long does it take to design a course?

For a focused course of 2 to 4 hours, plan one to two weeks of design before recording, then as long again for production. Design time is rarely wasted, because it removes lessons you would otherwise record and later delete.

Is ADDIE the only model?

No. ADDIE is the classic sequence, but many creators prefer backward design or quick iterative cycles: build a small version, run it with a pilot, fix and expand. Pick the model that makes you test with real learners early.