{"course":{"slug":"visual-storytelling","title":"Making It Say Something","url":"https://addaly.com/learn/visual-storytelling","level":"beginner","track":"create"},"lesson":{"slug":"showing-work-to-strangers","title":"Three tests, no budget, twenty minutes","url":"https://addaly.com/learn/visual-storytelling/showing-work-to-strangers","position":57,"of":64,"minutes":9,"answer":"Comprehension can only be measured on somebody who does not already know, so run a five-second test.","keyIdea":"Comprehension can only be measured on somebody who does not already know, so run a five-second test, an explain-back and a first-click on five strangers: hand it over in silence and never ask whether they like it."},"body":{"format":"markdown","text":"## You cannot test it on yourself, for a reason you already know\n\nYou know what the design says, so you will always be able to find the message in it. The curse of knowledge applies to pictures exactly as it applies to sentences. The only instrument available is somebody who does not already know.\n\nThe good news is that the tests that work are cheap, fast, and need no software.\n\n## The five-second test\n\nShow the design for five seconds. Take it away. Ask two questions:\n\n1. **What was that about?**\n2. **What did it want you to do?**\n\nThat is the whole test, and it directly measures the thing hierarchy exists for. If three people give three different answers, you do not have a hierarchy problem to investigate: you have the answer already.\n\nFive seconds is generous. A poster gets less, a feed gets a second and a half. Use five because it produces an answer rather than a shrug, and remember you are grading against a harder standard than the one you tested at.\n\n## The explain-back test\n\nFor anything longer (a video, a page, a brochure) hand it over, say nothing, and afterwards ask them to explain it to you as though you had not seen it.\n\nThe gap between their account and your intention is the measurement. Three specific gaps are common:\n\n- They get the subject and miss the point. The piece is clear and says nothing.\n- They repeat your words back without content. They have absorbed the phrasing and not the claim.\n- They fixate on a detail. Something in the piece is louder than the message, and now you know which thing.\n\n## The first-click test\n\nFor anything with an action in it: **where would you tap to do X?**\n\nThis one is worth knowing because the first click is unusually predictive. People who begin a task by going the right way finish it far more often than people who begin wrongly, and recovery from a wrong first move is much rarer than designers expect. If the first click is wrong, nothing downstream saves it.\n\n> Figure: Twenty minutes, five strangers, three tests. Steps: 1. Hand it over in silence (Explaining first installs the answer); 2. Five seconds (What was it about, what did it want); 3. Explain back (Say it to me as if I had not seen it); 4. First click (Where would you tap to do this). Never ask whether they like it. Liking is answered socially, and people say yes kindly about a design that told them nothing.\n\n## How to run it for nothing\n\n- A family member who is not in your field. The requirement is not expertise, it is ignorance of this particular project.\n- A neighbour, a shopkeeper, the person next to you in a queue. Ten seconds of somebody's time is not a big ask and most people are pleased to be asked.\n- A group chat: post the image, ask the two questions, do not explain first.\n- Public critique communities, which are free and mostly generous, provided you show the brief so people are critiquing against the right target.\n\nFive people is a good first batch. You will read that five testers find 85 per cent of problems, and that figure comes from a model assuming every problem is independent and roughly equally likely to be noticed. Real problems are not equally likely, and distinct groups of users hit different ones, so the honest version is: **five is enough to find the obvious failures, and not enough to conclude there are none.**\n\n## The two things that ruin a test\n\n**Explaining the design first.** The moment you say \"so this is about the workshop's new hours\", you have installed the answer in their head and the test is over. Hand it over in silence. This is much harder than it sounds and it is the single commonest way people waste a tester.\n\n**Asking whether they like it.** Liking is not a measurement and people answer it socially: they will say yes to be kind, and they will say it about a design that told them nothing. Ask what it said, what they would do, what they noticed first. Never ask what they think of it.\n\n## Watch the hands and the face\n\nWhat people do is more reliable than what they report. Somebody who pauses, squints, turns the page back, or scrolls past and returns has told you something they will not say out loud, often because they have already blamed themselves for not understanding.\n\nThat self-blame is worth knowing about: when a design fails, most people assume they were being slow. They will not report the failure, and it is still a failure.\n\n## The limitation\n\nThese tests measure comprehension, and comprehension is not the whole job. They will not tell you whether the piece is moving, whether it will be remembered next week, whether it suits the client's standing, or whether it will work in a market you have not sampled. Five friends in one city are not your audience.\n\nThey are also blunt about degree. They reliably distinguish *clear* from *unclear* and are poor at distinguishing *good* from *better*, which means they are for catching failures rather than for choosing between two strong options. For that you need the brief, a judgement, and a willingness to decide."},"check":{"question":"Why does explaining the design before a five-second test invalidate it, rather than merely helping the tester get started?","options":[{"text":"The explanation takes time, so the tester has effectively had longer than five seconds to absorb the design.","correct":false,"tempting":"Treats it as a timing error that a stricter stopwatch would fix, when the damage persists however briefly you speak."},{"text":"Testers become reluctant to contradict the designer once they know what was intended, so they under-report problems.","correct":false,"tempting":"Names a real social pressure, but the test would still fail with a tester who felt entirely free to disagree."},{"text":"Verbal explanation engages a different channel from visual comprehension, so the two interfere and produce a confused response.","correct":false,"tempting":"Invokes channel interference to describe what is simply prior knowledge, and would wrongly imply a written explanation is safe."},{"text":"It installs the answer you were trying to measure, so their response reports your words rather than the design's.","correct":true}]},"prev":{"slug":"reading-the-numbers","title":"Which number answers which question","url":"https://addaly.com/learn/visual-storytelling/reading-the-numbers"},"next":{"slug":"taking-feedback-without-drowning","title":"\"Make the logo bigger\" is a symptom, not an instruction","url":"https://addaly.com/learn/visual-storytelling/taking-feedback-without-drowning"},"authorship":"Founder and series editor: Dr Mustafa Mun. Written by AI models to a written standard. Exam keys checked blind by a second model: 3,665 of 3,666 agreed. Academic review invited.","review":{"status":"Draft, open for academic review"},"cite":"Addaly (2026) 'Three tests, no budget, twenty minutes', in Making It Say Something. Edited by M. Mun. Addaly. Available at: https://addaly.com/learn/visual-storytelling/showing-work-to-strangers (Accessed: 13 October 2026).","licence":{"name":"CC BY-NC-SA 4.0","url":"https://creativecommons.org/licenses/by-nc-sa/4.0/"},"reader":{"schema":1,"title":"Three tests, no budget, twenty minutes","courseTitle":"Making It Say Something","module":{"index":8,"slug":"judgement-and-the-body-of-work","title":"Judgement, and the work that accumulates"},"answer":null,"keyIdea":"Comprehension can only be measured on somebody who does not already know, so run a five-second test, an explain-back and a first-click on five strangers: hand it over in silence and never ask whether they like it.","hero":{"kind":"figure","figure":{"t":"figure","label":"Figure 8.1","figure":{"type":"flow","steps":[{"label":"Hand it over in silence","detail":"Explaining first installs the answer"},{"label":"Five seconds","detail":"What was it about, what did it want"},{"label":"Explain back","detail":"Say it to me as if I had not seen it"},{"label":"First click","detail":"Where would you tap to do this"}],"title":"Twenty minutes, five strangers, three tests","caption":"Never ask whether they like it. Liking is answered socially, and people say yes kindly about a design that told them nothing."},"alt":"Figure: Twenty minutes, five strangers, three tests. Steps: 1. Hand it over in silence (Explaining first installs the answer); 2. Five seconds (What was it about, what did it want); 3. Explain back (Say it to me as if I had not seen it); 4. First click (Where would you tap to do this). Never ask whether they like it. Liking is answered socially, and people say yes kindly about a design that told them nothing."}},"blocks":[{"t":"h2","id":"you-cannot-test-it-on-yourself-for-a-reason-you-already-know","runs":[{"text":"You cannot test it on yourself, for a reason you already know"}]},{"t":"p","runs":[{"text":"You know what the design says, so you will always be able to find the message in it. The curse of knowledge applies to pictures exactly as it applies to sentences. The only instrument available is somebody who does not already know."}]},{"t":"p","runs":[{"text":"The good news is that the tests that work are cheap, fast, and need no software."}]},{"t":"h2","id":"the-five-second-test","runs":[{"text":"The five-second test"}]},{"t":"p","runs":[{"text":"Show the design for five seconds. Take it away. Ask two questions:"}]},{"t":"ol","start":1,"items":[[{"text":"What was that about?","b":true}],[{"text":"What did it want you to do?","b":true}]]},{"t":"p","runs":[{"text":"That is the whole test, and it directly measures the thing hierarchy exists for. If three people give three different answers, you do not have a hierarchy problem to investigate: you have the answer already."}]},{"t":"p","runs":[{"text":"Five seconds is generous. A poster gets less, a feed gets a second and a half. Use five because it produces an answer rather than a shrug, and remember you are grading against a harder standard than the one you tested at."}]},{"t":"h2","id":"the-explain-back-test","runs":[{"text":"The explain-back test"}]},{"t":"p","runs":[{"text":"For anything longer (a video, a page, a brochure) hand it over, say nothing, and afterwards ask them to explain it to you as though you had not seen it."}]},{"t":"p","runs":[{"text":"The gap between their account and your intention is the measurement. Three specific gaps are common:"}]},{"t":"ul","items":[[{"text":"They get the subject and miss the point. The piece is clear and says nothing."}],[{"text":"They repeat your words back without content. They have absorbed the phrasing and not the claim."}],[{"text":"They fixate on a detail. Something in the piece is louder than the message, and now you know which thing."}]]},{"t":"h2","id":"the-first-click-test","runs":[{"text":"The first-click test"}]},{"t":"p","runs":[{"text":"For anything with an action in it: "},{"text":"where would you tap to do X?","b":true}]},{"t":"p","runs":[{"text":"This one is worth knowing because the first click is unusually predictive. People who begin a task by going the right way finish it far more often than people who begin wrongly, and recovery from a wrong first move is much rarer than designers expect. If the first click is wrong, nothing downstream saves it."}]},{"t":"h2","id":"how-to-run-it-for-nothing","runs":[{"text":"How to run it for nothing"}]},{"t":"ul","items":[[{"text":"A family member who is not in your field. The requirement is not expertise, it is ignorance of this particular project."}],[{"text":"A neighbour, a shopkeeper, the person next to you in a queue. Ten seconds of somebody’s time is not a big ask and most people are pleased to be asked."}],[{"text":"A group chat: post the image, ask the two questions, do not explain first."}],[{"text":"Public critique communities, which are free and mostly generous, provided you show the brief so people are critiquing against the right target."}]]},{"t":"p","runs":[{"text":"Five people is a good first batch. You will read that five testers find 85 per cent of problems, and that figure comes from a model assuming every problem is independent and roughly equally likely to be noticed. Real problems are not equally likely, and distinct groups of users hit different ones, so the honest version is: "},{"text":"five is enough to find the obvious failures, and not enough to conclude there are none.","b":true}]},{"t":"h2","id":"the-two-things-that-ruin-a-test","runs":[{"text":"The two things that ruin a test"}]},{"t":"p","runs":[{"text":"Explaining the design first.","b":true},{"text":" The moment you say “so this is about the workshop’s new hours”, you have installed the answer in their head and the test is over. Hand it over in silence. This is much harder than it sounds and it is the single commonest way people waste a tester."}]},{"t":"p","runs":[{"text":"Asking whether they like it.","b":true},{"text":" Liking is not a measurement and people answer it socially: they will say yes to be kind, and they will say it about a design that told them nothing. Ask what it said, what they would do, what they noticed first. Never ask what they think of it."}]},{"t":"h2","id":"watch-the-hands-and-the-face","runs":[{"text":"Watch the hands and the face"}]},{"t":"p","runs":[{"text":"What people do is more reliable than what they report. Somebody who pauses, squints, turns the page back, or scrolls past and returns has told you something they will not say out loud, often because they have already blamed themselves for not understanding."}]},{"t":"p","runs":[{"text":"That self-blame is worth knowing about: when a design fails, most people assume they were being slow. They will not report the failure, and it is still a failure."}]},{"t":"h2","id":"the-limitation","runs":[{"text":"The limitation"}],"role":"limit"},{"t":"p","runs":[{"text":"These tests measure comprehension, and comprehension is not the whole job. They will not tell you whether the piece is moving, whether it will be remembered next week, whether it suits the client’s standing, or whether it will work in a market you have not sampled. Five friends in one city are not your audience."}]},{"t":"p","runs":[{"text":"They are also blunt about degree. They reliably distinguish "},{"text":"clear","i":true},{"text":" from "},{"text":"unclear","i":true},{"text":" and are poor at distinguishing "},{"text":"good","i":true},{"text":" from "},{"text":"better","i":true},{"text":", which means they are for catching failures rather than for choosing between two strong options. For that you need the brief, a judgement, and a willingness to decide."}]}],"check":{"question":[{"text":"Why does explaining the design before a five-second test invalidate it, rather than merely helping the tester get started?"}],"options":[{"text":[{"text":"The explanation takes time, so the tester has effectively had longer than five seconds to absorb the design."}],"correct":false,"tempting":[{"text":"Treats it as a timing error that a stricter stopwatch would fix, when the damage persists however briefly you speak."}]},{"text":[{"text":"Testers become reluctant to contradict the designer once they know what was intended, so they under-report problems."}],"correct":false,"tempting":[{"text":"Names a real social pressure, but the test would still fail with a tester who felt entirely free to disagree."}]},{"text":[{"text":"Verbal explanation engages a different channel from visual comprehension, so the two interfere and produce a confused response."}],"correct":false,"tempting":[{"text":"Invokes channel interference to describe what is simply prior knowledge, and would wrongly imply a written explanation is safe."}]},{"text":[{"text":"It installs the answer you were trying to measure, so their response reports your words rather than the design’s."}],"correct":true,"tempting":null}],"order":[0,1,2,3],"why":null},"prev":{"slug":"reading-the-numbers","title":"Which number answers which question","minutes":9},"next":{"slug":"taking-feedback-without-drowning","title":"“Make the logo bigger” is a symptom, not an instruction","minutes":9}}}