How to Make a Quiz

The five steps below build one real quiz from scratch: a five question Workplace Safety Refresher that auto grades itself and shows each person their score. Every screenshot is the live editor, and you can take the finished quiz further down the page.

1. Name your quiz and pick a type

   Start a new quiz and give it a name the audience will recognise when it arrives in their inbox. "Workplace Safety Refresher, March" tells someone what it is and when it is from. "Quiz 1" does not.

   Then choose the quiz type. This one setting decides the rest of your build: a scored quiz gives you an answer key and a total, while a personality quiz swaps the answer key for outcomes. You can change the individual settings afterwards, so pick the closest match and move on.

   

   

   You should now have an empty quiz with a name and a type, ready for its first question.

2. Add your questions

   Add your first question and choose a question type. Multiple choice is the right default for most knowledge checks. True or false is quick but carries a fifty percent guess rate, so use it sparingly. Short answer tests real recall, but expect to mark some responses by hand unless you list every accepted spelling.

   Write the question as a complete sentence and include only what someone needs to answer it. Keep your answer options similar in length and style, because an option that is noticeably longer or more detailed than the others reads as the correct one even to someone who does not know the material.

   

   

   You should now see your question in the list, with its answer options, and an empty slot for the next one.

   Which question type to use

   Choosing a question type

   Type	Best for	Watch out for
   Multiple choice	Most knowledge checks. Lets you use common misconceptions as the wrong options.	Accidental clues, such as one option being longer or using absolutes like "always".
   True or false	Quick checks and clear policy statements.	A fifty percent guess rate, and wording that is arguable either way.
   Short answer	Real recall, with no options to jog the memory.	Spelling variants and manual marking.
   Matching	Terms and definitions, or sorting things into groups.	Too many pairs turns it into a reading test.
   Scenario question	Judgement and decisions, which is what most workplace tests actually need.	Long set ups packed with detail that does not matter.

3. Mark the correct answers and set the points

   This is the step that turns a list of questions into an assessment, and it is one click per question: select the correct option and it becomes the answer key. Every response is then graded automatically, without you writing a single rule.

   This is not universal. In some form builders, scoring means writing a logic rule for every answer, one at a time, then a second set of rules to decide what each total means. If you have ever abandoned a quiz halfway through building it, that is usually why.

   Leave every question worth the same points unless you have a reason not to. If some questions genuinely matter more, such as the safety critical ones or the items tied to a certification requirement, weight those higher and keep the rest even.

   

   You should now see a correct answer marked on every question, and a points value against each one.

4. Decide what people see at the end

   A quiz that just says "7 out of 10" teaches nobody anything. Before you publish, decide what the last screen does.

   If the quiz is for learning, show the score and the correct answers, with a sentence explaining why each one is right. That explanation is where the learning actually happens, and it costs you a few minutes to write.

   If the quiz is a real assessment, or you plan to reuse it, hold the answers back. Show the score, and keep the answer key private so the quiz survives being shared around.

   If a result triggers something, such as a certificate or a follow up, set the pass mark now and keep it in proportion to the difficulty and the consequences of failing.

   If you are collecting email addresses, the ask interacts with length. In our benchmark, short quizzes of five to twelve questions completed 86.9 percent of the time without an email ask and 76.8 percent with one; at thirteen questions or more the effect reversed, 67.6 against 72.1. The interruption hurts a short quiz, while a long one has already earned the commitment. If capturing leads is the whole point, using a quiz to capture leads is a different build with its own trade offs.

   

   Do this before you publish

   Add up the total points, check the pass mark against it, then take the quiz twice yourself: once answering everything correctly, and once getting a few wrong on purpose. It takes two minutes and it catches almost every scoring mistake before your audience does.

   You should now know exactly what a person sees the moment they finish.

5. Preview it, then publish and share

   Preview the quiz as a taker rather than as its author. You are checking three things: that the questions read clearly when you are not the person who wrote them, that nothing is cut off on a phone, and that the result screen says what you expected in step four.

   Then publish. You get a link you can paste into an email, a chat or a class group, and you can embed the quiz directly in a page so people never have to leave your site.

   Where the free account comes in

   Worth knowing up front, because most guides are vague about it. Building and saving a quiz here needs no account at all, answer key included. Creating a free account is what unlocks the shareable link and the response reporting. So you can do the whole job and see exactly what you have made before deciding whether to sign up.

   

   

   You now have a live quiz on a URL, and responses will appear as people complete it.

After you send it: reading the results

Publishing is not the end of the job. A quiz beats a slide deck because it tells you what people did not know, and that only shows up once answers arrive.

Two numbers matter. The score distribution tells you whether the quiz was pitched right: everyone scoring full marks means it was too easy to be informative, and everyone failing usually means unclear rather than hard. The per question breakdown is the valuable one. A question most people get wrong is either a real gap in what they know, which tells you what to teach next, or a badly worded question, which tells you what to fix.

You should now be able to see who has answered, how they scored, and which questions caused the most trouble.

Reporting is where a quiz stops being a one off. If you want to go further than this page covers, there are more step by step guides for the rest of the product.

Back to top