Sefism early access is open for X and Instagram followers and university students.Get early access
Learning and skills

When should I stop taking courses and start building things?

Tauseef Fayyaz

Answered by Tauseef Fayyaz

Asked 11 timesUpdated 18 Sept 2026
Short answer

Sooner than feels comfortable. Once you know the basics of a language and one tool, start building, and let the project decide what you learn next. Courses become reference material, used for an hour to unblock a specific problem, not a path you finish before you are allowed to start.

What people tell me

I have completed several online courses and certificates, and I understand the material while I am watching. But when I sit down to build my own project, I do not know where to start, so I sign up for another course to feel more prepared. It has been months like this. I feel productive but I have nothing of my own to show.

A composite of the messages behind this question, with personal details left out.

Key takeaways

  • Courses feel like progress because they are comfortable. Building feels hard because it is where the learning actually happens.
  • Start building once you can write basic programs in one language and use one framework at a beginner level.
  • Flip the ratio: roughly 20 percent learning material, 80 percent building.
  • Use courses as reference: watch one lesson to solve a specific problem, then go back to the project.
  • Being stuck while building is not a sign you are unready. It is the part that makes you better.

Courses are comfortable, building is uncomfortable

This is one of the most common patterns I see, and it is completely understandable. A course gives you a clear next step, a sense of progress and a certificate at the end. Building your own project gives you a blank screen and a lot of not knowing what to do.

So people keep doing the comfortable thing and call it preparation. But the discomfort of building is not a sign you are unready. It is exactly the skill employers are paying for: taking an unclear problem and working it out.

The honest threshold

You are ready to start building once you can do these:

  1. Write basic programs in one language: variables, loops, functions, working with lists and objects.
  2. Follow a beginner tutorial in one framework and understand roughly what each part does.
  3. Use Git to save and push your work.

That is it. You do not need to finish the advanced course, the second certificate or the full bootcamp. If you can do those three things, you are ready, and every extra course before building has diminishing returns.

Flip the ratio

Most people stuck in courses spend something like 90 percent of their time consuming and 10 percent creating. Try the reverse for a month:

ActivityShare of your time
Building your own projectAbout 80 percent
Watching or reading learning materialAbout 20 percent, only for what the project needs now

The project decides what you learn next. Need a login? Learn authentication. Need to save data? Learn that database. Learning with a specific problem in front of you sticks far better than learning in case you might need it.

Use courses as reference, not a path

You do not need to stop using courses. Change how you use them. Instead of watching a course from start to finish, treat it like a manual:

My app needs to upload images. I will watch the one lesson on file uploads, then go and implement it in my project.

One targeted lesson, then straight back to your own code.

Pick a first project that is small and real

Do not start with a clone of a huge product. Pick something small, useful and yours:

  • A tracker for something you actually do: expenses, study hours, workouts.
  • A tool that solves a small problem for your family, friends or a local business.
  • A simple app built around a public API you find interesting.

Then build it in slices. Get one feature working end to end before adding the next. A tiny working app beats an ambitious broken one. If you want ideas at a larger scale, the project ideas library has plenty.

What to do when you get stuck

You will get stuck a lot. That is normal. A useful routine:

  1. Read the error message properly, top to bottom.
  2. Search for the exact error.
  3. Try to reduce the problem to the smallest example that fails.
  4. Ask AI or a person, but ask it to explain rather than write the whole thing.
  5. Write down what fixed it.

Each time you get unstuck on your own project, you learn more than a whole course module could teach you, because you had to find the answer yourself.

How you know it is working

After a month of building, you should notice: you search for specific things instead of watching whole courses, you can explain the choices in your project, and you have something you can show someone. That last point matters most. A working project you built, however small, says more about you than any number of certificates.

An example week of building

Here is what a realistic first week might look like for someone who has finished a JavaScript course and wants to build a study tracker:

DayWhat you do
MondayWrite down five features. Cross out three. Keep: add a study session, see this week's total
TuesdaySet up the project and Git. Make a page with a form that logs a session to the screen
WednesdaySave sessions so they survive a refresh. Watch one short lesson on local storage when stuck
ThursdayShow the weekly total. Fix the bug where dates are off by a day
FridayTidy the code, write a short README, push it and deploy it for free

By the weekend, you have a small working app you built yourself, and you have learned about storage, dates and deployment because the project needed them. That is more real progress than a week of course videos.

A self-check to catch yourself slipping back

Every Sunday, answer these:

  1. How many hours did I spend building versus watching?
  2. Did I push code this week?
  3. Did I sign up for anything new this week, and did my project actually need it?
  4. Can I show someone one thing that works now but did not last week?

If you notice yourself reaching for a new course, ask one more question: "What specific problem in my project will this solve?" If you cannot name one, close the tab and open your editor instead.

If you are still stuck

Read What to Do When You Are Stuck Learning to Code and How to Use AI as a Companion, Not a Crutch.

If you want someone to look at your specific situation, join Sefism and, as a member, book a 1:1 session. You can also browse the other career questions people have asked.

Was this answer helpful?

Ask your own

Your situation is not quite this one?

Members get written answers to their own questions, a roadmap built for them and feedback on their projects. Early access is open to X and Instagram followers and university students. Prefer to talk? A free call works too.

Work with me

Stuck on something specific?

Writing only gets you so far. If you want an answer to your situation rather than the general case, book a session and we will work through it together. Sessions are free for approved Sefism members, and a few slots open each week.

More questions people ask

All questions