A test series on a YouTube channel is one paper per video: a short test on the chapter you just taught, published free, with the link in the description and pinned in the top comment. A lecture proves you can teach. A paper proves your viewer can perform - and the second proof is the one that turns a subscriber into a student. The only cost is the time it takes to set, because MockSetu is free and takes no card.
This article is about the loop, not the paper itself. For how long a chapter test should be and which questions belong in it, read how to create a chapter-wise test online. For the same idea run by an institute with a building rather than a channel, see a free test series as a lead magnet, and for the version of this problem an audience on Telegram has, from PDF dumps to real mock tests.
Why the Paper Does Work the Video Cannot
A view tells you somebody watched. It does not tell you - or them - whether the explanation landed. A paper does both. The student finds out what they can actually do against a clock, and you find out which part of your own explanation did not survive contact with a question, per question, across everyone who sat it.
That matters most for the viewer who has never paid anybody. From the outside one confident voice over a whiteboard looks much like another. A question paper is checkable: set it well and the viewer has judged you for themselves, with nobody selling anything.
The friction is deliberately low. A visitor who is not signed in can open a published paper and sit the whole thing; signing in is asked for at the finish, so the results can be saved. A link under a video does not have to survive an account wall before question one.
Start With One Test That Matches One Video
The tempting first move is a full-length mock. It takes far longer to set and it tests a syllabus the viewer has not finished. Start with a paper shaped like the lecture: same chapter, same question types, sat in one sitting straight after the video.
- Pick a video that already does well. The paper rides on traffic that exists, not on traffic you are hoping for.
- Set it on that chapter only. One section is usually enough - add a second only when the chapter is really two topics with different question types.
- Take the marking scheme from the bulletin of the exam your viewers are preparing for, not from habit. On a chapter test a penalty discourages the very guesses that would have shown you where the confusion sits.
- Fill in the name, category, description and general instruction properly. All four are required before the exam exists at all, and all four are read by strangers.
- Open it from the dashboard menu - the item is called "Sit it as a student" - and read it the way a viewer will. A creator preview records nothing, so nothing you do there pollutes the results.
- Publish it. Only then share it.
Publish First: What the Description Link Actually Is
The Share action on your creator dashboard copies the exam's intro address to the clipboard, and it refuses outright on a draft. Publish first, or nothing is copied and you paste whatever was already on the clipboard into a description that goes out under a published video. Publishing needs at least one section with questions in it, so a paper you have only sketched cannot be shared by accident.
What the link opens is the instructions page, not question one. It reads in two steps - what to do before you begin, then the paper's own instructions - and carries the description you wrote plus a table of the paper built from live exam data at view time: one row per section, its questions, the maximum marks where a marking scheme is set, and each section's clock where sections are sat one at a time. Write that description for a stranger who arrived cold, not for your regulars.
Put the link in the description and in a pinned comment both. They sit in different parts of the page and not every viewer opens both, and the cost of covering both is one extra paste.
The Library Is a Second Doorway, and the Byline Is Yours
Your description link reaches people who already found you. The public exam library reaches people who did not. Every published exam gets its own card there, filterable by category, with your byline on it - the unique user ID you chose when you set the account up, which allows letters, numbers and underscores. If your channel handle is free in that format, take it, so the card and the channel read as the same person.
One detail should change how you name papers. The library's search box matches an exam's name and its category, plus the previous-year label on papers tagged as such. It does not search your description. A paper called "Test 4" can only be found by somebody already scrolling; the same paper named for its chapter and its exam can be found by somebody typing. The name is the searchable part, so spend a minute on it.
Two things on that card are not yours to switch on. The Mock / Previous Year tag, which puts a year badge on previous-year reconstructions, is a per-creator grant enforced on the server - a creator without it does not see the field at all and the paper reads as a mock. The Verified Creator badge beside a byline is granted on request, never automatically.
A lecture asks a viewer to believe you. A question paper lets them check for themselves - and a viewer who has checked does not need to be sold anything.
The Next Video Is Where the Loop Closes
This is the part that compounds, and the part that is easiest to skip. Once people have sat the paper, its analytics page hands you the next video's script. There is a per-question breakdown: how many got it right, how many got it wrong, how many left it unanswered, the average time spent on it across everyone who attempted it, and which wrong option was picked most often. Above that table sit three panels - Most Skipped, Most Reviewed and Common Misconceptions - each listing the top five questions by that measure. Your own attempts are filtered out of your exam's analytics, which is correct and briefly surprising the first time you look.
- Open Common Misconceptions and take the wrong option picked most often. Teach the distractor, not only the right answer - the question is interesting precisely because that option looked right.
- Read Most Skipped as a difficulty signal with a caveat: a question near the end of a section may be skipped for the clock rather than for the content. Check where it sat before you conclude anything.
- Most Reviewed is what people marked for review while they sat the paper. That is doubt rather than failure, and it tends to earn a clarifying line in the solution video rather than a full re-teach.
- Quote the figures in aggregate on camera and never with a name attached. The analytics give you counts, not a list of who answered what; the only names on the page are the usernames on the top-three leaderboard.
- Close the solution video with the next paper's link. That is the loop - video, paper, solutions driven by the data, next paper - and it is the part of this that gets easier each time.
Notice what this gives you that a comments section cannot. Comments come from the viewers willing to post. The analytics cover every attempt that was saved, the quiet student included, and they tell you which half of your explanation to repeat rather than which half flattered you.
Public Is the Point, Not a Leak
Published papers are public. Anyone with the link can sit yours, a rival channel included, and there is no private delivery to one batch, no paywall and no payments of any kind. Educators arriving from a coaching background read that as a hole in the product.
For a channel it is the distribution. A paper only your subscribers could open would be worth less to you, not more - the stranger who finds it through the library or a forwarded link is exactly the person the channel exists to reach. What you are protecting was never the questions. It is the byline on them.
What This Will Not Do for Your Channel
Stated plainly, because everyone else in this market oversells. There is no proctoring of any kind - no webcam, no lockdown browser, no tab-switch detection - so a published paper is a shop window and not an examination hall. There are no notifications about your tests: MockSetu sends transactional account mail for signup and password reset and nothing else, so announcing a new paper is your channel's job, not the platform's. There is no CSV or Excel export of results and no per-student report card. No white-labelling, no custom domain, no mobile app, no certificates, no attempt limits, no question shuffling, no LMS or Google Classroom integration, and no subjective or essay grading. If one of those is a hard requirement for you, this is a good time to find out.
The Second Paper Is the Series
One paper is a sample. A run of papers under one byline is a test series, and the series is what a viewer subscribes for. The cost drops sharply after the first one. Duplicating an exam carries the marking scheme, the timing groups and the language links into the copy, which arrives as a fresh draft with an empty leaderboard, so the next paper costs you the questions and little else - the copy is best-effort, so open it and check the marks before you build on it.
If the questions already exist as a PDF, the bulk route is MockSetu's published extraction prompt plus a JSON upload, documented in the JSON upload guide. Server-side PDF import exists but is off by default and switched on per creator on request, so plan around the prompt. Either way the importer leaves numeric, TITA and match-the-column questions for you to enter by hand, and the AI only extracts from a PDF you supply - it never writes questions from a syllabus.
Everything a creator can build here - sections with their own clocks, per-question marking, bilingual papers, a listing in the public library - sits behind one free account. Record the video. Set the paper. Pin the link. Then read what the paper tells you, and make that the next video.