I am constantly studying but feel like I am making no progress. How can I tell whether I am actually improving?

Busy and better are not the same thing

This question usually comes from people who are working hard. That is what makes it painful. The hours are real. The problem is that a lot of study time goes into activities that feel like learning but mostly build recognition: you see an idea explained and it makes sense. Recognition is not the same as being able to produce the idea yourself on a blank page.

So the first step is to stop measuring effort and start measuring ability.

What real progress looks like

Ask yourself whether, compared with a month ago, you can do more of these without help:

• Start a small project from an empty folder and get it working.
• Read an error message and find the cause yourself.
• Solve an unfamiliar problem of a type you have practised.
• Explain a concept, such as how a request reaches your server, in plain words.
• Read someone else's code and say what it does.
• Estimate roughly how long something will take and be close.

If the answer to most of these is "not really", your study method is the problem, not your ability.

Audit how you spend study time

Track one normal week and put each session into one of two columns:

A healthy split is something like one third passive, two thirds active. Most people who feel stuck are at nine tenths passive.

Set monthly checkpoints

Once a month, give yourself a short test that you cannot fake:

1. Build test. A tiny app in two to three hours with no tutorial, for example a to-do list with a backend, or a page that fetches and filters data.
2. Problem test. Five problems you have not seen, timed, no hints.
3. Debug test. Take one of your old projects, break something deliberately or pick an open bug, and fix it.
4. Explain test. Write a one-page explanation of something you studied this month, as if teaching a junior.

Write down the results. Next month, repeat with similar difficulty. That comparison is your real progress chart.

Redo old work cold

One of the simplest and most encouraging tests: pick a problem or small project from four to six weeks ago and redo it with no help. If it feels easier, faster or cleaner, you have improved, even if new material still feels hard. New material always feels hard. That is what makes it new.

Keep a learning log

Every day, one or two lines: what I did, what I got stuck on, what I understood. It takes two minutes. After two months, read it from the start. People are consistently surprised by how far they have come, because the change was too gradual to feel.

Signs your method needs to change

• You finish courses but cannot build anything similar without them.
• You recognise solutions instantly but cannot produce them.
• You rely on AI for most of the code and cannot explain parts of it.
• You keep switching topics before going deep on any.

If two or more apply, keep the same hours but move most of them into active work. You will feel slower for a couple of weeks. Then you will actually be faster.

A worked example of a monthly checkpoint

Say you have spent a month learning backend development with Node.js. A useful checkpoint on the last Saturday of the month might be:

• Build (two hours): an API with three endpoints for notes: create, list, delete, stored in a real database. No tutorial open.
• Debug (thirty minutes): rename a column in the database and fix everything that breaks, using only the error messages.
• Explain (twenty minutes): write half a page on what happens between a browser request and your database query.
• Problems (one hour): five easy problems you have not seen.

Record the time each took, where you needed to look something up, and what you could not finish. Next month, run a checkpoint of the same size on the next topic. If you finished more, looked things up less, or your explanation was clearer, that is real progress, whatever your feelings say.

A weekly self-check in three questions

Every Sunday, answer in writing:

1. What can I do now, without help, that I could not do last Sunday?
2. Where did I get stuck, and did I get unstuck myself or did someone (or AI) do it for me?
3. What was the ratio of building to watching this week?

If the answer to the first question is "nothing" for two weeks running, change the method before adding more hours. If the underlying problem is too much to cover, where to begin when there is too much to learn covers how to cut it down.

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. The resources library has practice material for the monthly checkpoints. If you want someone to look at your specific situation and review your progress honestly, join Sefism and, as a member, book a 1:1 session.