Chapter 21. The First 30 Seconds, 3 Minutes, 30 Minutes, and Day

Consider some hypothetical numbers. “Our game's tutorial completion rate is 80 percent.” When that number appears in a report, the room brightens for a moment. Eight out of ten people followed it to the end, so the tutorial seems well made. But if next-day return for the same game is 5 percent, those two numbers are not telling the same story. People followed the tutorial successfully but did not return the next day. We managed to make them finish something, but it did not give them a reason to come back. Getting someone to the end and getting them to return are different jobs. This is not the first time we have seen that mismatch. Chapter 2 paired an 80 percent completion rate with 20 percent next-day return. The numbers differ here, but the shape of the mismatch is the same. Now we will lay that mismatch out across time and find where the two paths diverge.

Chapter 20 fixed the boundaries of the first experience with two points: its beginning and its end. Two points alone, however, cannot tell us what happens between them. We still do not know whether the trip from beginning to end lasts 30 seconds or 30 minutes, nor when to give something and when to hold it back. This chapter divides that territory by time: the first 30 seconds, 30 seconds to 3 minutes, 3 to 10 minutes, 10 to 30 minutes, and one day. The user's state of mind changes in every interval, so what we should give them changes as well. The same explanation can be poison in the first 30 seconds and medicine at the three-minute mark. A first experience is as much about when we give something as what we give. One clarification: the clock in this chapter starts at first launch. The time before that, in ads and the store, belongs to the promise design discussed in Chapter 18.

A Person Who Asks a Different Question Every 30 Seconds

Let us return to our hypothetical game: the mobile game where people collect characters, talk with them, and watch short videos. Suppose someone has just seen an ad, installed the app, and opened it for the first time. We will follow what passes through their mind, 30 seconds at a time.

The first 30 seconds. They ask, “What is this? Is it for me? Is it safe?” They are not yet ready to do anything. While checking whether this screen matches the ad, they look around with the wariness of someone entering an unfamiliar place. If these 30 seconds are consumed by a long company logo, an unskippable intro, and a procession of consent forms, they forget what they came expecting. Between 30 seconds and 3 minutes, their guard relaxes a little. They ask, “What am I supposed to do?” and press something for the first time. If it responds, they feel reassured; if it does not, their finger stops. Between 3 and 10 minutes, once the controls feel familiar, they ask, “So what makes this fun?” They receive their first goal and either achieve it or get stuck for the first time. Between 10 and 30 minutes, if they have tasted the first pleasure, they ask, “Is this worth continuing? Does it have depth? Is there anything I can choose?” Then they close the app. The next day, whether they open it again is the real test of the first experience.

It is the same person, but the question changes every 30 seconds. Showing “Does it have depth?” in the first 30 seconds is too early. Keeping them stuck on “Is it safe?” after ten minutes is too late. Time design means answering in step with the changing rhythm of those questions.

Something should be made clear in advance. Numbers such as 30 seconds and 3 minutes are not laws of nature. The scale expands and contracts with genre and channel. Thirty seconds in a game whose rounds end in under a minute does not carry the same weight as 30 seconds in a game designed for long immersion. What remains constant is not the number, but the order of the questions: What is this? What do I do? What makes it fun? Is it worth continuing? Will I come back? Every genre climbs this ladder in the same order. A timetable merely sets that ladder to the rhythm of our game. Accordingly, when we read a leaking funnel, we should ask not “At what minute did they leave?” but “Which question went unanswered?”

The rhythm is faster and more distinct in MEJE Aidong World. Fans drop in briefly before sleep or during a break, and if an Aidong does not appear and react early, they close the app at once. Aidong World's first 30 seconds therefore offer no explanation. They offer one Aidong whose feel comes from K-pop idols: cute, and responsive to touch. Within the first few minutes, the fan makes that Aidong their favorite, then quickly gives it a name and begins caring for it. Guidance designed for a long-session user is pulled forward into a shorter rhythm for a fan dropping by in a spare moment. The scale has been compressed, but the order of the questions remains unchanged. Aidong World's quick rhythm is the first proof of the principle stated above.

The First 30 Seconds Are About Identity and Safety

If you cannot establish safety in 30 seconds, your 30 minutes of content do not exist.

The first 30 seconds carry the most weight, and this is where most people split away. Yet what we commonly do in those 30 seconds is the exact opposite of what they need. Wanting to show our best work first, we play the intro video we labored over most, explain the world, and try to teach the controls. The user in the first 30 seconds, however, is not ready to receive any of that. They want only three things: to understand at a glance what this is (identity), to confirm that it matches what they saw in the ad (expectation match), and to feel reassured that it is not somewhere suspicious (trust).

Identity means showing at a glance what kind of game this is. If the first screen shows combat, people read it as a combat game; if a character greets them, they read it as a character game. Expectation match means that the first screen does not contradict the promise made by the ad or store page. If we promised cute collecting and brought them in only to show a grim battle on the first screen, they feel deceived. Trust loosens their wariness toward something unfamiliar, because safety matters more than fun on the first screen. If an unfamiliar character is too aggressive, if the first button leads to payment, or if the app demands personal information immediately, trust breaks at the outset. Fun later rarely repairs trust broken that early. Do not try to teach anything in the first 30 seconds. Reassure people and meet their expectations. Teaching comes after they open up.

From 30 Seconds to 3 Minutes: The First Input and Response

Once their guard relaxes, the user presses something for the first time. Our job in this interval is to give them a reason to press and make the world respond to what they pressed. This is where the feedback from Chapter 15 goes to work. Return a signal that the input registered and a signal that it had meaning, both quickly.

Tutorials divide into two approaches here. One explains and instructs: a finger icon points to a spot and says, “Tap here,” and the user follows directions. The other lets them learn by bumping into things: give no explanation, let them touch, and allow them to discover by pressing. Which approach is right depends on the user. A follow-the-instructions tutorial is fast and safe, but it seats the user in the passenger seat. They spend their first three minutes feeling that they did what they were told, not that they did something themselves. Learning by collision puts them in the driver's seat, but risks letting them get lost. Most experiences therefore combine the two. Point out just one or two things they absolutely need to know, then let them discover the rest by touching. Ask only one thing at a time, and teach even that through action rather than explanation. The user did not come to spend their first three minutes reading.

From 3 to 30 Minutes: Achievement, the Big Picture, and Choice

The first goal arrives between 3 and 10 minutes. Give one small, clear objective and let the user complete it—or get stuck once and then break through. This is the ending point fixed in Chapter 20, the interval in which the first core pleasure arrives. Whether they complete their first character, hold their first conversation, or win their first round, this is when the user first feels, “Ah, this is good.” As Chapter 17 explained, failure in this interval should be rhythm, not punishment. The small curve of getting stuck and then breaking through makes the achievement feel alive.

Between 10 and 30 minutes, the user's question changes again: “Is this worth continuing?” Having seen the first pleasure, they now judge its depth. Give them a glimpse of growth, the big picture, and choice. Show a little of what will open later—do not show everything, because that is a spoiler—and give the user something to choose according to their tastes. This is where T9 from Chapter 3, first choice and social connection, goes to work. They customize a character, give it their colors, and feel for the first time, “This is mine.” A pile of choices in the first 30 seconds is a burden, but choices after the first pleasure are freedom. Timing turns the very same options into either burden or freedom.

And then, one day. Whether the user opens the app again the day after closing it is the final test of the first experience. This is T10 from Chapter 3, first return. Strictly speaking, the choice beyond ten minutes (T9) and the return (T10) fall outside the FTUE boundary drawn in Chapter 20. They belong to the NUX that follows. Why, then, does this chapter's table include columns for 30 minutes and one day? Because of the handoff. The score is recorded beyond the boundary, but the seed that earns that score must be planted inside it. So we ask: did the first session plant an excuse to return? Whether they miss the puppy they made yesterday or wonder about something left unfinished, the first session must contain one reason to call them back.

▶ Three Things to Apply to Your Screen

Does the first 30 seconds answer “What is this, and is it safe?” or spend those 30 seconds on logos, an intro, and terms?

Are you charging people an explanation before they see the first pleasure even once? Do you let them act first and postpone the explanation?

Does the first session plant one “reason to return tomorrow”? If even one answer is no, people leak from that interval.

Where Does It Leak Most, and Is a Long Tutorial Kind?

The real reason to divide time into intervals is to pinpoint where people leak out. Chapter 3 divided the funnel by thresholds; here we divide it by time. A substantial share of people who download a free game never return. Designers who work with free games often say next-day return tends to settle somewhere between 20 and 40 percent. In other words, more than half disappear within a day. But that single number does not tell us what to fix. People do not all leave at the same point. They leave from different places in each interval. We must look not only at how much the whole experience leaks, but at which interval has the steepest drop.

Spread this across some hypothetical numbers. Suppose 100 people open the first screen, 70 remain after 30 seconds, 50 after 3 minutes, 35 after 10 minutes, 25 after 30 minutes, and about 8 return the next day. The distribution shows which interval is steepest. If the largest leak is the first 30 seconds, where 100 falls to 70, people left before they saw any fun. The problem is not fun but identity and trust. Conversely, if 25 make it to 30 minutes but only 8 return the next day, the first session worked but failed to plant a reason to come back. The repair handle changes completely depending on where the steepest interval lies. If the first 30 seconds leak, shorten the intro. If the next day leaks, plant an excuse to return.

One gaming convention now faces its final checkpoint: the belief that “a long tutorial is kind, and an NPC should explain everything.” This is the third checkpoint aimed at “a lot all at once.” Chapter 16 examined the noise of reading, Chapter 19 examined an overload of tasks, and here we examine the advance charge on time.

First, consider what this convention assumes. Games are complex enough to have many controls, rules, and systems to learn. Games have therefore long placed a friendly guide on the screen. An NPC at the village entrance approaches, teaches the controls, points through the menus, and explains the systems step by step. The longer and more thorough this guidance, the kinder it was thought to be. Creators and gamers alike learned that a well-made game explains everything without leaving anything out.

An ordinary person does not receive this kindness as kindness. The time they bring to the first screen may be no longer than a single subway stop or the brief wait for an elevator. If an NPC approaches someone who drifted in to fill that spare moment and launches into a five-screen explanation, the game charges their time before giving them a single taste of fun. By the second explanation screen, their thumb is already over the Home button. YouTube plays a video as soon as they tap. Short-form feeds deliver the next item as soon as they swipe. To someone trained by that pace, “Listen to everything first, then begin” is not kindness but a toll. A screen where an NPC explains everything feels considerate to the speaker but burdensome to the listener. A gamer may read the long explanation as “care”; a newcomer reads it as “delay.”

The fee this convention charges an ordinary person is time paid before fun, and the bill arrives during the most expensive first 30 seconds. The longer the time to value, the more people leak away along the way. A long tutorial does not merely lower its completion rate; it sends people away before completion. This explains the mismatch between the 80 percent tutorial completion rate and 5 percent next-day return above. The 80 percent who reached the end are people who listened to the explanation, not necessarily people who had fun. Completion rate measures “Did they hear my entire explanation?” It cannot measure “Did they come to like my game?”

Mobile game developers sometimes call this the “onboarding cliff.” Users follow a well-constructed tutorial to the end, only to be dropped without guidance into a suddenly complex main game the instant it finishes. We believed we had taught them everything with care, yet they leave in a rush precisely where the teaching ends. The cliff signals that following to the end and coming to like something are two different stories.

Keep the reassurance that guidance provides; discard the inertia that says kindness requires explaining everything. Let people act first and explain later. Point out the one thing they must know at that moment, then reveal the rest little by little where it becomes necessary. Do not pile the entire explanation onto one screen. Place a small cue where they first encounter each function. Onboarding offers several patterns we can borrow: a tiny tooltip that briefly explains an element at first encounter, a task list that shows what to do next, a progress bar that shows how far they have come, and an empty state whose text provides guidance when there is nothing there yet. Empty states are especially common in first experiences. A blank screen with no collected characters, friends, or history greets a new user. Fill that absence with a line such as “Your puppy will soon appear here,” and the empty screen becomes guidance toward the next action instead of a void. Let explanations trickle out when the user needs them instead of pouring everything out at once. That replaces the long tutorial. A reward at the tutorial's end follows the same logic: if someone spent time learning, do not leave them empty-handed. Let one small achievement wait at the end of the lesson.

Therefore, Decide One Different Priority for Each Interval

Once divided by time, first-experience design becomes much clearer. For each interval, decide separately: “What question is the user asking here, so what should we give and what should we postpone?” In the first 30 seconds, give identity and safety while postponing explanation. Up to 3 minutes, give the first input and response while postponing depth. Up to 10 minutes, give the first achievement. Up to 30 minutes, give the big picture and choice without revealing everything. One day later, let one reason to return go to work. When one priority has been set for each interval, there is no need to argue over where every element belongs.

This design leads into measurement. Once the intervals are defined, we can see how many people remain in each one, and the interval with the steepest drop becomes the next place to repair. Remember one trap, however. A high tutorial completion rate does not mean the first experience succeeded. Completion rate measures only “Did they do everything we told them to do?” The numbers that matter are the share who reached the first pleasure and the share who returned the next day. So establish one operating rule: never report tutorial completion rate by itself. If that number enters a report, place first-pleasure reach rate and next-day return beside it and report only the three-line set. A number that becomes a boast when it stands alone finally becomes a diagnosis when it stands with the other two. We must look at numbers that reveal not what we made people finish, but what we made them like. Choosing those numbers and learning how to read them is the work of the next part.

Reference Content

Examples in other media that reveal the concepts in this chapter.

Video Games

Half-Life's opening tram ride: For the first few minutes, the game does not take control of the screen. It simply lets players look around from inside the tram, establishing identity and atmosphere first, then postponing control instructions until after they disembark.
An unskippable procession of logos at boot: This charges time before showing any fun, filling the first screen with a toll instead of identity and safety.
A mobile game that drops players into the main game without guidance as soon as the tutorial ends: A counterexample that demonstrates the “onboarding cliff,” where users leave in a rush exactly where the teaching ends.
An RPG opening that begins with five screens of NPC explanation: An example of a long tutorial that charges time before the user tastes the fun.

App UX

Slack's empty channels and Slackbot: An empty screen with no messages yet explains what the channel is for and guides the next action. Slackbot speaks to the user and gets them to reply, letting them learn to send messages by doing it.

The remaining examples are collected in the “Chapter 21 Appendix” at the end of this text (grouped as Appendix D in the book).

Design Note ▶ Try It Yourself

Divide our game's first experience into five boxes. In each box, write one thing to give and one thing to postpone.

First 30 seconds: give (　　　) / postpone (　　　) 30 seconds–3 minutes: give (　　　) / postpone (　　　) 3–10 minutes: give (　　　) / postpone (　　　) 10–30 minutes: give (　　　) / postpone (　　　) One day later: reason to return (　　　)

When all five are filled, inspect the “postpone” entry in each box. If a long explanation or NPC guidance appears under “give” in the first-30-seconds or 3-minute box, rewrite it. Move the explanation to the interval in which the user first needs that function. Leave only identity and safety in the first-30-seconds box.

Finally, add one more line: “If the number we are currently watching is tutorial completion rate, write first-pleasure reach rate and next-day return beside it.” See whether the three numbers tell the same story or contradict one another. (The detailed blank table for what to give and postpone in each interval, along with a sheet for expected drop-off by interval, appears in Appendix B.)

In one line: A first experience is as much about when we give something as what we give. The first 30 seconds are for identity and safety; up to 3 minutes, the first input and response; up to 10 minutes, the first achievement; up to 30 minutes, the big picture and choice; and one day later, a reason to return. The user's question changes in every interval, so what we give and postpone must change too. The convention that “a long tutorial is kind, and the NPC should explain everything” charges an ordinary person's most expensive first minute as a time fee. Let them act first and explain later, and watch first-pleasure reach and return instead of tutorial completion. Next chapter: This completes the engineering of the first impression: what to foreshadow, how to resolve collisions, where to set the boundary, and what to give at each moment. This is the end of Part IV. Yet every time we said, “We need to see where it leaks,” we postponed one question. What numbers are we actually looking at? What are we trying to increase? Part V begins by building that dashboard.

Chapter 21 Appendix: Reference Content Collection

This collection shows how other media practice time design: dividing what to give and what to postpone according to the ladder of questions users climb at 30 seconds, 3 minutes, 30 minutes, and one day—What is this? What do I do? What makes it fun? Is it worth continuing? Will I come back? The examples mix games and apps with film, television, music, novels, webtoons, short-form content, and offline experiences. Each ends with a “What to examine” prompt. The length of the scale expands or contracts by medium, but the order of the questions remains the same. Read each example as a test of that claim.

Video Games

Super Mario Bros. World 1-1 (Nintendo, 1985): The first screen leads Mario naturally toward a Goomba, letting the player collide with the rule “jump on it to defeat it” without words, then reveals rules one at a time through progress. What to examine: This is the standard for embedding the first 30 seconds of learning in action without explanation. Does the first lesson on our first screen arrive as text or action?
Mega Man X's intro stage (Capcom, 1993): During play, it drops the player into a pit they cannot escape normally, leading them to discover wall jumping without a single line of explanation. What to examine: This is a carefully shaped path for learning by collision. Compare it with the chapter's two approaches: does a control discovered independently remain longer than one performed on command?
Bayonetta's opening action (PlatinumGames, 2009): The first scene immediately hands players a spectacular battle that is difficult to die in, establishing identity and tactile pleasure first. The controls tutorial covering punches, kicks, and dodges comes afterward. What to examine: Its sequence follows the chapter's claim exactly—the first 30 seconds belong to identity, not learning.
Portal's first test chamber: Environmental cues such as wall color and lighting guide the eye toward where to fire a portal, allowing players to discover the solution through experiment rather than written instruction. What to examine: When the environment replaces explanation, reading time becomes action time. Which of our instructions could be moved into environmental cues?
Animal Crossing's daily loop: Newly buried fossils, replenished rock resources, and villagers to greet are tied to the real-time clock, creating a reason to launch again tomorrow even after stopping today. What to examine: This plants the “reason to return one day later” not in an individual piece of content but in the system's time structure. What bait for tomorrow does our first session leave behind?

App UX

Duolingo's first entry: It postpones account registration and immediately gives the newcomer a short translation or multiple-choice exercise. Learning itself becomes the first experience instead of an explanation, and sign-up comes afterward. What to examine: The flow answers the first three minutes' question—“What do I do?”—with action. Are our first three minutes filled with reading or doing?
Candy Crush Saga's early levels: The first few rounds are made so easy that they are almost impossible to lose, handing the player an early “I'm good at this” achievement before gradually increasing difficulty. What to examine: It guarantees the first achievement in the 3-to-10-minute interval through design rather than luck. Does our first achievement reach everyone?
Wordle's one puzzle a day: This word puzzle, which spread at the end of 2021, gives everyone the same single problem each day. Once it is solved, there is nothing more until tomorrow. The restraint of offering no more becomes a reason to return, while a one-line colored-grid result drove word of mouth. What to examine: It fills the “one day” box with a restriction rather than added content. A reason to return does not have to be a reward.

Device UX

Tamagotchi care alerts: When it grows hungry or its mood falls, a beep calls the owner back to care for it. This short-cycle summons creates a reason to return after stepping away. What to examine: The device itself manufactures and offers the excuse to return. Where is the line between a welcome call and an annoyance?

Short Form / Streaming

The first one-to-three-second hook in short-form video: Among TikTok and Reels creators, the assumption that viewers immediately swipe away if the first second or two fails to capture attention functions almost like standard grammar. Platform production guides likewise recommend a hook in the opening seconds. What to examine: In this medium, the gateway is the first 3 seconds rather than 30. Even as the scale shrinks by channel, the order of questions—starting with “What is this?”—remains unchanged.
Netflix's “Skip Intro” button: Since 2017, series openings have offered a skip button that removes the time spent repeatedly watching an introduction viewers already know. What to examine: The same intro establishes identity in episode one but becomes a toll by episode five. This shows that timing determines value as much as content.

Live-Action Film

Film's three-act structure: The same information can be dull in the introduction and tense in the middle, so the structure designs when to give it. What to examine: The value of information comes from its position as much as its content. Imagine moving one of our explanations to a different interval.
The opening hook and slow buildup: If the first few minutes fail to hold viewers, later masterpieces never get the chance to be seen. What to examine: This is the film version of the funnel—the good things later exist only if the beginning is passed.

Live-Action TV / Drama

Breaking Bad's pilot cold open: It first grabs viewers with a man in his underwear driving an RV through the desert and aiming a gun at approaching sirens. Only afterward does it return to ordinary life and explain who he is and how he got there. What to examine: This reverses the order by placing the hook first and explanation later. Does our worldbuilding explanation come before our hook?
The end-of-episode cliffhanger: It plants an excuse at the end of the first session that calls the audience onward and makes them start again. What to examine: Since the reason to return is planted at the end of a session, what does the last screen of our first session leave behind?

Music

The buildup from intro to verse to chorus: The opening seconds catch the ear, and the first reward—the chorus—arrives on a precise beat. What to examine: The structure promises the arrival time of the first reward through rhythm. At what minute have we promised ours?
Intros shortened around streaming's 30-second accounting threshold: Because a stream must play for a certain amount of time to count, long intros and late choruses became costly, reportedly shortening intros and pulling hooks forward. What to examine: Measurement structures can reshape the content's time design in reverse. How are our metrics carving our opening?

Literature

The pacing of a novel's first chapter: It places a hook on the first page and postpones background explanation until the reader is absorbed. What to examine: The distribution of explanation is the craft of Chapter One. Recount the amount of explanation on our first screen.
Chapter One of Gone Girl (Gillian Flynn, 2012): It drops the reader into the moment a wife disappears on the morning of her fifth wedding anniversary, laying down anxiety first, then revealing the couple's history later through alternating diaries and memories. What to examine: This is the restraint of holding a setting without explaining it immediately. Which pieces of our worldbuilding can wait until a later session?
An information-heavy first chapter that dumps the setting in the opening: It loses the reader by spoon-feeding explanation before they can enter the story. What to examine: This is an advance charge on time in writing. Compare it with the volume of our tutorial's first screen.

Newspapers / Articles

The inverted pyramid: This longstanding news-writing convention puts the most important fact in the first sentence, then adds progressively less important detail, so the reader retains the core no matter where they stop. What to examine: The structure assumes departure. Is our first experience designed so that anyone who leaves at any interval still carries away one clear identity?

Webtoons

The opening of Tower of God: It first grabs readers with one question—“What do you desire?”—and the lure of the Tower, then gradually reveals the Tower's rules and world through tests. What to examine: One question carries the reader through 30 seconds while the setting travels inside action. What action carries our world explanation?
The fast opening of Episode One: The first few panels of a vertical scroll make readers want to continue, postponing world explanation. What to examine: This convention pulls the scale forward to fit the medium's pace. Does our scale fit our channel's pace?

Web Novels

A fast opening that prevents Episode One drop-off: The convention is well established that if the beginning of the first installment does not catch readers, the later story will never be shown. What to examine: A serialized medium stakes everything on the funnel's first interval. Compare its choice with our answer about what belongs there.

Real-World Work Procedures

A new hire's first day, week, and month: The same information is poison if poured out on day one and medicine if given in the week when it becomes necessary. What to examine: The ladder of questions appears unchanged at work—Where am I? What do I do? Am I useful? Move the chapter's table for dividing what to give and postpone into onboarding.
Job training that lets people try first and explains later: It allows someone to complete one small task before making them read the entire manual. What to examine: This is the workplace version of action before explanation. On which screen does the first action appear in our tutorial?

Offline / Everyday Life

A gym's first consultation and session: Experienced trainers commonly advise against pushing a newcomer to their limit on day one, because several days of soreness may drive them away. Instead, end the first session lightly and finish with one small success. What to examine: This is the offline version of the “one day” box, where first-day intensity determines next-day return. How far does our first session push people?