Playtest survey builder for games
This builder writes a short playtest survey for your game: 8 to 12 ordered questions sized to your session, each with an answer type and a time to ask it. Pick the playtest type, genre, session length and goals. Copy the survey as plain text or Markdown, or download a CSV, then paste it into Google Forms or Tally.
Your playtest survey
10 questions, about 7 minutes. Target for this session length: 6 to 10 questions.
The survey takes one question for each goal in turn until the target is full.
Thank you for playing our game. You are one of the first people to play it. This survey has 10 questions and takes about 7 minutes. We test the game, not you. There are no wrong answers, and honest answers help us most, also when they are negative.
Consent: when you send your answers, you agree that we use them to improve the game. We do not ask for your name. Please do not write personal data in the text boxes.
Right after the session
Put these in the form that testers open when the session ends.
- How often do you play games like this one?Single choice
- Most weeks
- A few times a year
- I tried one or two
- Never before this test
- Overall, how was the game?Scale 1 to 51 = very bad, 5 = very good
- Which moment do you remember best? Tell us what happened.Open text
- Where were you bored or annoyed? Tell us what happened.Open text
- When the session ended, how much did you want to keep playing?Scale 1 to 51 = not at all, 5 = very much
- How would you describe this game to a friend in one sentence?Open text
- How easy or hard was it to understand what to do in the first minutes?Scale 1 to 51 = very hard, 5 = very easy
- Describe a moment when you did not know what to do next, if there was one.Open text
- In your own words, what is the goal of the game?Open text
- What else do you want to tell us?Open text
How to paste this into Google Forms or Tally
Neither tool reads this CSV as a list of questions, as far as we know. Add the questions by hand: it takes a few minutes for a survey of this size. Keep the CSV as your own record of what you asked in each playtest.
Google Forms
- Create a blank form. Paste the intro and the consent line into the form description.
- Add one question for each row. Use "Linear scale" from 1 to 5 for a scale and type the two labels. Use "Multiple choice" for single choice, "Checkboxes" for multi choice and "Paragraph" for open text.
- Add each option of a choice question. Use "Copy options" in this tool to get them one per line.
- For a tester ID, use More, then "Pre-fill form": fill in the ID field, click "Get link" and send each tester their own link.
Help pages: Google Docs Editors Help: Choose a type of question for your form, Google Docs Editors Help: Publish and share your form with responders ("Send a form with pre-filled answers").
Tally
- Create a blank form. Paste the intro and the consent line as text at the top.
- Type / to add one input block for each row: "Linear scale" for a scale (set the steps and the labels in the block options), "Multiple choice" for single choice, "Checkboxes" for multi choice and "Long answer" for open text.
- Click "Copy options" in this tool and paste the lines into the first option of the block. Tally adds one option for each line.
- For a tester ID, type /hidden to add a hidden field named tester_id, then add ?tester_id= and the ID to the link you send to each tester.
Help pages: Tally Help: Input blocks, Tally Help: Bulk insert option lists, Tally Help: Hidden fields.
How to use it
- Pick the "Playtest type", the "Game genre" and the "Session length".
- Tick the goals under "Goals: what do you want to learn?". Untick them all to get the standard survey.
- Optional: type the "Name of your game", and tick "Add a tester ID question, to join answers with analytics data".
- Under "Your playtest survey", read the question count, the minutes and the note. If a goal got no question, pick fewer goals.
- Click "Copy as plain text", "Copy as Markdown" or "Download CSV", then follow "How to paste this into Google Forms or Tally".
How it works
The builder picks from a bank of 54 questions that we wrote. Each question has a goal, an answer type (a scale from 1 to 5 with both ends named, single choice, multi choice or open text) and a time to ask it: during the session, right after it, or a day later. Questions made for one genre or one playtest type only appear for that genre or type.
Every survey has the same frame: an optional tester ID, then "How often do you play games like this one?" and "Overall, how was the game?", and at the end the open question "What else do you want to tell us?". The frame asks broad questions before specific ones, which games user researcher Steve Bromley recommends.
The target count depends on the session length: 8 questions for under 15 minutes, 10 questions for 15 to 45 minutes, 12 questions for 45 minutes to 2 hours, 12 questions for several sessions over days. These targets are our practice, not a standard, and the tool shows them as a range from 6 questions to the target. The builder fills the space between the frame questions one goal at a time, in a fixed order (fun, onboarding, difficulty, controls, bugs, intent, price), until the target is full. With no goal ticked, it uses the core questions that work for most playtests.
An in-person session adds up to 2 neutral prompts to ask aloud during play. A demo, a closed test, a Steam Playtest or a test over several sessions can add up to 2 questions for a message one day later. The time is our estimate: 20 seconds for a closed question and 60 seconds for an open one, rounded to whole minutes.
questions = frame (2, or 3 with a tester ID) + goal questions + 1 closing question
target = 8 (under 15 minutes), 10 (15 to 45 minutes), 12 (45 minutes to 2 hours), 12 (several sessions over days)
minutes = round((closed questions × 20 + open questions × 60) ÷ 60), at least 1Worked example
Example with made-up inputs: a Steam Playtest of a roguelike, 15 to 45 minute sessions
A studio runs a Steam Playtest of a made-up roguelike called Ember Rogue. It picks "15 to 45 minutes", ticks onboarding, difficulty and willingness to wishlist or buy, and adds the tester ID question. The target is 10. The frame takes 4 questions (tester ID, play habits, overall rating, closing question), which leaves 6 for the goals: 2 each.
The builder gives 10 questions. Difficulty gets the roguelike question "When you failed, how fair or unfair did it feel?", and intent gets "Is this game on your Steam wishlist now?", which only appears for a demo, a closed test or a Steam Playtest. 7 questions are closed and 3 are open, so the time is 7 × 20 + 3 × 60 = 320 seconds, about 5 minutes.
How to read the result
Count the survey against its target. The survey for a session under 15 minutes takes about 5 minutes at 8 questions, and the longest one about 8 minutes at 12. If the note says a goal got no question, the session is too short for all your goals: drop one, or run a second playtest for it.
With 5 to 10 testers, read each answer as a story, not as an average. Jakob Nielsen found that tests with 5 users find about 85% of usability problems, so a small round is enough to find the big ones. To compare ratings between builds, the consumer playtest method from Microsoft Game Studios used 25 to 35 players per test.
On the difficulty question the middle of the scale is the target, not the top. Read it next to how far each tester got. For intent, trust the facts (a wishlist added, the build played again) over any rating.
Our practice for open answers: group them by the moment or feature they name, and count how many testers mention the same one. A problem that 3 of 5 testers name is worth fixing before the next round.
Decide what you want to learn first
Write down two or three things you need to know about this build before you tick any goal, for example "do new players understand the first run" or "is the second boss too hard". These are your research objectives, and Steve Bromley advises to set them before you write questions. Tick only the goals that serve them. A survey is good at measuring opinions and simple facts. To learn why a tester got stuck, watch them play or look at where they quit in your play data.
Why the survey is short
We keep the survey short, so that testers still have energy for the open questions. "Keep the survey short" is also one of the tips of games user researcher Steve Bromley. If you pick more goals than a short session can cover, the builder tells you which goals got no question. The target counts and the time estimate (20 seconds for a closed question, 60 seconds for an open one) are our own practice.
What the survey does not ask
No question asks a tester to predict a purchase, and no question asks the tester to design a fix. Games user researcher Steve Bromley advises to never ask about future behaviour, because it correlates poorly with what players really do, and to ask about things that players did before. So the questions about buying ask for the last similar game the tester bought, for a wishlist that exists today and for the actions after the test.
When to ask each question
Most questions belong right after the session, when the memory is fresh. An in-person session adds up to 2 neutral prompts to ask aloud during play. A demo, a closed test, a Steam Playtest or a test over several sessions adds up to 2 follow-up questions for a message one day later. Steam does not collect playtest feedback for you: Valve says that collecting feedback is up to you, for example in a forum, on Discord or on your own site.
Pilot the survey before you send it
Before the playtest, have one teammate or friend fill in the form while you sit next to them, and ask what they think each question means. Five minutes of this catches confusing words and broken answer options. Steve Bromley lists this pilot as a step of its own. Then keep the questions and the scales the same from build to build, so the answers stay comparable.
Where to put the survey link
Put the link where testers see it at the end of play: on the last screen of the build, in the pause menu, or in the message that sends the build. For a Steam Playtest, Valve suggests a Steam forum, Discord or your own site for feedback, and you can post the link in a community announcement. With the tester ID question, give each tester a link with the ID filled in, as the paste notes above explain for Google Forms and Tally.
Limits
The builder writes the questions only. It does not host a form, collect answers or analyse them. Paste the survey into Google Forms, Tally or another form tool.
Estimate (20 seconds per closed question, 60 per open one) for the minutes. The count and the minutes also include prompts asked aloud during play and questions sent a day later, which are not part of the same form.
The question bank is fixed and in English. You cannot edit, reorder or add questions in the tool: copy the survey and edit it in your form tool.
Price answers are opinions. Read them as a range that testers accept, not as a sales forecast. No survey answer predicts how many players will buy.
A survey tells you what testers felt and remember, not why they did something. Watch testers play, or compare their answers with their play data, to find the cause.
Frequently asked questions
How many questions should a playtest survey have?
We aim for 8 questions after a session under 15 minutes and up to 12 after a long one. By our estimate, that takes about 5 to 8 minutes. This is our practice, not a standard. A short survey with open "why" questions gives better answers than a long form, because testers rush the end of a long one.
What questions should I ask in a playtest survey?
Start broad: how often the tester plays games like yours and an overall rating. Then ask about your goals, such as the first minutes, difficulty or controls, with both ends of each scale named, for example "easy or hard". Add open questions about the best and the worst moment. Ask about past actions, not future plans, and end with an open question.
How many playtesters do I need?
To find the big problems, a small round is enough: Jakob Nielsen found that 5 users find about 85% of usability problems. Fix them and test again. To compare ratings between two builds with some confidence, you need more: the consumer playtest method from Microsoft Game Studios used 25 to 35 players per test. Recruit players of your genre, not only friends.
Does Steam Playtest have a built-in feedback survey?
No. Valve says that collecting feedback from players is up to you. Steam handles signups and access to the playtest build, but not questions or answers. Make the survey in a form tool, then share the link in the build, in a Steam community announcement, in a Steam forum or on Discord. A tester ID lets you join the answers with play data.
Should I ask testers if they would buy my game?
No. Answers about future behaviour correlate poorly with what players really do, as games user researcher Steve Bromley notes. Ask about facts instead: is the game on their wishlist now, which similar game they bought last and why, and, a day later, what they did since the test. Those answers say more about real interest than a promise to buy.
Why is there a tester ID question?
It is optional. With an ID, you can put the answers of a tester next to what that tester did in the build, for example in Manalens after identify() with the same ID. Without an ID, you can read the answers but cannot join them to play data. Send each tester a link with the ID filled in.
Sources
- Steve Bromley, "How to write a playtest survey", Games User Research, 12 January 2022., read
- Valve, Steamworks documentation: Steam Playtest ("Collecting feedback from players is up to you")., read
- Jakob Nielsen, "Why You Only Need to Test with 5 Users", Nielsen Norman Group, 18 March 2000., read
- John P. Davis, Keith Steury and Randy Pagulayan, "A survey method for assessing perceptions of a game: The consumer playtest in game design", Game Studies 5(1), October 2005., read
- Google Docs Editors Help: Choose a type of question for your form., read
- Google Docs Editors Help: Publish and share your form with responders ("Send a form with pre-filled answers")., read
- Tally Help: Input blocks., read
- Tally Help: Bulk insert option lists., read
- Tally Help: Hidden fields., read
Last reviewed . The survey is built in your browser. We do not send your choices anywhere.
Next step
Call identify() with the same tester ID in your build, and Manalens shows each tester's profile with their event feed and activity, so you can read the survey answers next to what that tester did. See player profiles →