How to design an app SO addictive people cant stop using it — Transcript
Full transcript
- 0:00Duolingo, TikTok, and Tinder all have
- 0:02one thing in common, and it's the reason
- 0:05why people keep using them every single
- 0:07day without even thinking about it.
- 0:09Because it doesn't really matter how
- 0:10good your app looks, if people stop
- 0:13using it on the first day, then all of
- 0:15your work goes to waste. Now I have
- 0:18designed some of the most popular apps
- 0:20on the App Store, and so I have seen
- 0:22which psychological levers you need to
- 0:24pull to keep people coming back every
- 0:27single day. And so in this video I'm
- 0:30going to show you the science behind why
- 0:32these apps are so addictive that people
- 0:35can't stop using them. All of these apps
- 0:37use the same six-step framework, and it
- 0:40starts before the user even opens the
- 0:42app and goes all the way to them coming
- 0:44back every single day. So let's go
- 0:47through each of the steps together, and
- 0:49for each one I'm going to tell you how
- 0:50they do it and how you can do it on your
- 0:53own app. Every single app struggles to
- 0:55make users open it. The average person
- 0:57has 93 apps on their phone, but on any
- 1:00given month they only open about half of
- 1:02them. The rest of them just sit there.
- 1:05But think about the last time you opened
- 1:07TikTok. It was probably not through a
- 1:09notification. You were just bored, you
- 1:11felt the urge, and before you could even
- 1:13think about it, your thumb was already
- 1:15over the icon. That urge is what I call
- 1:18the itch. [music]
- 1:20So how does an app turn into an itch?
- 1:23There's this great book called Hooked
- 1:24that separates the reasons for opening
- 1:26apps into two categories. First, there's
- 1:29the external triggers, which are things
- 1:31like notifications or red badges on your
- 1:34app icon. And second, there's internal
- 1:37triggers, which are feelings. For
- 1:39example, you might feel bored or curious
- 1:41or feel like you're behind on your
- 1:43Spanish. And most of the time internal
- 1:47triggers are what's going to make you
- 1:48open an app. So when you're bored, you
- 1:51open TikTok.
- 1:52But let's look at the contrary example.
- 1:54Let's say you downloaded a meditation
- 1:56app in January, and on day one you
- 1:59actually meditated for minutes, you felt
- 2:01great, and it was good. But by day
- 2:03three, you forgot that the app was even
- 2:06there. Now, forgetting about the app
- 2:08doesn't mean that you stopped feeling
- 2:10stressed or whatever other internal
- 2:12trigger that made you download a
- 2:15meditation app. It just means that the
- 2:17app failed to associate itself with that
- 2:20feeling. And I would bet that your own
- 2:23app is making the same mistake.
- 2:25Go look at the things that are written
- 2:27on your app and the notifications you
- 2:28send people. Are they generic things
- 2:31like we miss you or are they speaking to
- 2:34the feeling of the user? The best apps
- 2:37start with external triggers and over
- 2:39time associate themselves with the
- 2:42feeling. That's why you don't need a
- 2:43notification to open TikTok. Boredom is
- 2:46the notification. Here's how to apply
- 2:48the itch to your app. First, write down
- 2:51the feeling that your app answers. It
- 2:53can be boredom, curiosity, or
- 2:56excitement. Then, go write down the
- 2:59notifications that you can send people
- 3:01that speak to that feeling. For example,
- 3:04dating apps tell you someone liked you.
- 3:06They don't tell you keep using our app.
- 3:09That's corporate speak. Nobody cares.
- 3:11And last but not least, make sure that
- 3:14you know when your users feel that
- 3:16feeling and that you can remind them
- 3:18that your app exists when that feeling
- 3:20arises. For example, if someone feels
- 3:23bored on their commute and your app is
- 3:25about boredom, make sure to remind them
- 3:27that the app exists during their
- 3:30commute. But the itch only gets people
- 3:32to open your app. What happens next
- 3:34decides if they start using it. And that
- 3:37is what I'm calling the lazy thumb.
- 3:40BJ Fogg runs a behavioral design lab at
- 3:42Stanford, and he has a very simple way
- 3:44of explaining this. Your motivation to
- 3:46do something goes up and down during the
- 3:48day, and you can't really control it.
- 3:51Now, what you can control is how easy it
- 3:54is to take action. So, instead of
- 3:56waiting for your users to be highly
- 3:58motivated to do something, just give
- 4:00them something easier to do so that even
- 4:03when their motivation is lower, they can
- 4:04simply do it. Let's say you're trying to
- 4:06track what you eat. You just had lunch
- 4:09and you want to log it. When I broke
- 4:11down the calorie tracking app Yazio on
- 4:13episode 60 of the app breakdown, I
- 4:15couldn't even log my food from the home
- 4:18screen on just one or two taps. I had to
- 4:20like scroll, go into a menu, and then
- 4:22type things.
- 4:23Now compare that with their competitor
- 4:26Cal AI, where you take a photo of your
- 4:28plate and it's done. It's the whole
- 4:31product. They do exactly the same thing.
- 4:34They accomplish the same mission, but
- 4:36one does it in multi-step and the other
- 4:38one does it with a single photo. As a
- 4:41result, the easier one to do got over 15
- 4:44million downloads in two years. And
- 4:46notice how all of the most addictive
- 4:48apps make the core action extremely
- 4:51simple. TikTok, one swipe. Tinder, one
- 4:54swipe. Duolingo, one tap in a gigantic
- 4:58button at the bottom of the screen to
- 5:00start a new session. To apply this to
- 5:02your own app, first write down what the
- 5:04core action of your app is. Then count
- 5:08how many steps it takes to do it and cut
- 5:10it to one and be very ruthless. Once you
- 5:14know what that action is, make the
- 5:15button to perform it the biggest thing
- 5:18on the screen and ideally place it at
- 5:20the bottom where the thumb is because
- 5:22half of the people going through your
- 5:24app are probably going to be using the
- 5:26phone with one hand. So just make it
- 5:28reachable. And third, lower the bar for
- 5:31what counts as done. For example,
- 5:33Duolingo used to only count a streak if
- 5:35you completed your full daily goal. And
- 5:38then they moved on to counting the
- 5:40streak even if you do just one lesson.
- 5:43And as a result, retention went up 3.3%.
- 5:47But sometimes even one tap is too many
- 5:49if it's too hard to decide what to tap
- 5:52on.
- 5:53[music]
- 5:54You'd think people want to choose for
- 5:55themselves, right? But more often than
- 5:58not, they just go with whatever is
- 5:59already picked for them. The most famous
- 6:01example of this is a study done in 2003
- 6:04around organ donation in Europe. In some
- 6:07countries, you have to choose to donate
- 6:10your organs. And, as a result, only
- 6:12about 23% of people donate their organs.
- 6:16But, in other countries, you have to
- 6:18choose not to donate your organs. And,
- 6:21as a result, about 86% of the population
- 6:25donates their organs. So, despite being
- 6:28the same decision, there's this gigantic
- 6:30discrepancy in what people choose to do,
- 6:33simply because in some countries they
- 6:35have one paved path, and in other
- 6:37countries they have the other paved
- 6:39path. Now, going out of those paths
- 6:42takes work, and so people simply don't
- 6:45do it.
- 6:46Now, if people take the paved path on
- 6:48something as important as organ
- 6:50donation, you can imagine how much
- 6:52effort they're going to put into leaving
- 6:54the paved path on simple things like
- 6:56choosing which button to press on an
- 6:58app. And that's why TikTok never asks
- 7:01you what you want to watch. It's already
- 7:03playing. Duolingo doesn't give you a
- 7:05myriad of options. It just shows you the
- 7:07next lesson and lights it up. And Tinder
- 7:10doesn't ask you who you want to see
- 7:12next. It just shows you someone. That is
- 7:15the paved path, picking the next option
- 7:17for your user and not making them think.
- 7:20So, how can you pave the path of your
- 7:22own app? First, think about the action
- 7:24that you wrote on the lazy thumb and
- 7:26make that the first thing that people
- 7:29see when they open the app. When a user
- 7:31opens your app, they should not see a
- 7:33menu or a bunch of options. They should
- 7:35see the one option that matters. Then,
- 7:38browse your app and list out everywhere
- 7:41that you make users take a decision. And
- 7:44for every decision that you make them
- 7:45take, make sure that there's a default
- 7:48option. This default option should be
- 7:50what's better for you and your user, and
- 7:52you're going to see that more often than
- 7:54not people are going to take that
- 7:56default option. So, these three tricks
- 7:58are how addictive apps get you in, an
- 8:01itch to make you open the app, a very
- 8:03easy first action to get you to do
- 8:05something, and then a path that is
- 8:07decided for you. But getting you in is
- 8:09only half the battle, and now they have
- 8:12to hook you. And this is where the next
- 8:14three points come in, backed by TikTok's
- 8:17own internal documents. In 2024, a
- 8:20lawsuit in Kentucky made some of
- 8:22TikTok's internal documents public by
- 8:24mistake. And in those documents, TikTok
- 8:27has a name for the moment that they get
- 8:29you hooked. They call it the habit
- 8:31moment, and it happens exactly after you
- 8:34watch 260 videos. Now, how do they get
- 8:38you to watch so many videos so fast? The
- 8:41reason why this works goes back to a
- 8:43scientist called B. F. Skinner in the
- 8:451950s and his pigeons. So, what he did
- 8:49is he separated two groups of pigeons
- 8:51and gave each one a button. When they
- 8:53peck that button, they get food. Now,
- 8:56the difference is for the first group,
- 8:58when they peck the button, they always
- 9:00get food. For the second group, when
- 9:02they peck the button, they don't know if
- 9:04they get food or not. It's random. So,
- 9:06he ran this experiment for a while and
- 9:08after some time simply stopped giving
- 9:11food to the pigeons. In the first group
- 9:13of pigeons, the ones that always used to
- 9:14get food, they stopped pretty fast
- 9:17because they were not getting food
- 9:18anymore. But the second group simply
- 9:20kept going for a long time because they
- 9:23never knew when they were going to get
- 9:25food. It was this variable reward that
- 9:28kept them going exactly like a slot
- 9:30machine. And now, here's the crazy part.
- 9:33Tinder's own co-founder has said that
- 9:35this study inspired the swipe on Tinder.
- 9:39Every swipe on TikTok is like pulling a
- 9:41lever on a slot machine. And that's why
- 9:43I call the fourth point the slot
- 9:45machine.
- 9:47Now, think about the way the average app
- 9:49does rewards. Let's say it's a workout
- 9:51app. You do a workout, you get a badge.
- 9:53You think it's amazing. It's your first
- 9:55badge. Then you do another workout, you
- 9:57get a second badge. You don't really
- 9:59care as much. And by the 10th badge, you
- 10:01know exactly what's coming and so your
- 10:03brain has stopped caring.
- 10:06That's the slot machine. Make make
- 10:08unpredictable and vary when they come
- 10:11and how large they are.
- 10:13To implement this, simply think what the
- 10:15reward of your app actually is and then
- 10:19make it unpredictable. Make it so that
- 10:21users don't actually know what to
- 10:23expect, but give it to them fairly
- 10:25often. But unpredictable rewards only
- 10:28work if there's always another one
- 10:30coming. And who knows better about
- 10:32keeping things coming than As a Raskin,
- 10:35the designer that invented the infinite
- 10:37scroll and who says that by his own
- 10:39estimate 100,000 human lives are wasted
- 10:43scrolling every single day. That's the
- 10:46power of removing the end of a page. And
- 10:49that's why I call 0.5 of the addictive
- 10:51framework the bottomless pit.
- 10:55Most things in life come with stopping
- 10:57cues, like the end of the day, the end
- 10:59of a YouTube video, or the bottom of a
- 11:01page. And those are moments where your
- 11:03brain asks, should I keep going or is
- 11:06this the time to stop?
- 11:08If you remove all of these cues, then
- 11:10the user themselves has to make a choice
- 11:14to leave. And as we saw on the paved
- 11:16path, people usually don't make their
- 11:18own choices. And so they simply stay
- 11:21when you remove those cues. And most
- 11:23apps, without even realizing it, are
- 11:25designed to make users stop using them.
- 11:29Every empty search screen, every dead
- 11:31end, every single place where you don't
- 11:34lead the user somewhere else is an
- 11:36opportunity for their brain to go,
- 11:38should I stop now? By contrast,
- 11:41addicting apps never show you a way out.
- 11:43On Duolingo, when you finish a lesson,
- 11:45the biggest button is to start another
- 11:47one. On TikTok, there's no end to the
- 11:49feed. On Tinder, there's no end to the
- 11:51stack. On Netflix, the next episode
- 11:54starts playing before you even decided
- 11:56if you want to watch it or not. And
- 11:58that's the bottomless pit, always giving
- 12:01your user what's next, the next action,
- 12:03the next episode, and never showing them
- 12:05a way out. So if you want to apply this
- 12:07to your own app, start by listing every
- 12:10single dead end, the end of lists, empty
- 12:13search results, homes with error states,
- 12:16list everything that does not lead the
- 12:19user to somewhere else. And then simply
- 12:22take that list and make it lead users
- 12:24somewhere. If it's an empty search
- 12:26result, show them something that they
- 12:28might be looking for. If it's the end of
- 12:30a list, show them something else that
- 12:32they might be interested in. If it's the
- 12:34end of an episode, show them whatever is
- 12:36next, but never tell them that it's
- 12:39over. And now that you have your users
- 12:41engaged and doing multiple actions, you
- 12:43just need to make sure that they come
- 12:44back. And that's where point six of the
- 12:47addictive framework comes in. Netflix
- 12:49estimates that their recommendation
- 12:50algorithm saves them a billion dollars a
- 12:53year. Why? Because every show that you
- 12:56watch on Netflix teaches them what you
- 12:59like to watch. And so by the time you've
- 13:02watched a bunch of things, canceling
- 13:03Netflix feels like starting over. You
- 13:06don't want to explain again what you
- 13:08like to watch. And most addictive apps
- 13:10work in the same way. Every time you use
- 13:13them, you leave a little bit of yourself
- 13:15behind. You leave your taste, your
- 13:18playlists, your connections, your
- 13:20friends. And so it feels like a piggy
- 13:23bank. Every time you use them, you put a
- 13:25little coin in, and the more you use
- 13:27them, the harder it's going to be to
- 13:29simply walk away. That's why the sixth
- 13:31and last point of the framework is
- 13:33called the piggy bank.
- 13:35It's when an app asks you to put in a
- 13:37little work, your time, your
- 13:39preferences, your data. And so it
- 13:41becomes more personalized and you become
- 13:43more tied to it. Imagine starting TikTok
- 13:46over on a new account. Your current feed
- 13:49learned from every single video that you
- 13:51watched. And so starting over feels like
- 13:54you're leaving some part of yourself
- 13:56behind. And every time someone uses an
- 13:59addictive app, the app gets more
- 14:02personalized to them. So look at your
- 14:04own app. If someone uninstalls it today,
- 14:07will they actually lose anything? And if
- 14:09the answer is no, users will not come
- 14:13back. How do you fix this? First, you
- 14:15start tracking what your users do on the
- 14:17app. Every search that they make, every
- 14:19button that they click, every piece of
- 14:21information that your users give you,
- 14:23treat it as very valuable because it
- 14:26teaches you what they're actually coming
- 14:28to your app for. And that you should
- 14:30just give them more of. But first, you
- 14:32need to track them. And then secondly,
- 14:35after you know what they're looking for,
- 14:38give them a way to build that up. Give
- 14:40them progress, give them levels, give
- 14:42them an algorithm that becomes better.
- 14:44The options are very much limitless, but
- 14:47give them something that they can build
- 14:49on every day and that they are very
- 14:51aware that they can lose if they never
- 14:54come back. So now you know the science
- 14:56behind what makes these apps addictive,
- 14:58but it's one thing to hear about it and
- 14:59it's another one to see how it works
- 15:01inside the real app. So inside of this
- 15:04video right here, I break down Finch,
- 15:06the self-care app that's actually
- 15:08addictive so that you see exactly how
- 15:11they pull users in and how you can apply
- 15:13it to your own app.
About this transcript
This page contains the full transcript of How to design an app SO addictive people cant stop using it by App Breakdown, generated from the public captions YouTube serves with the video. The transcript has 2,762 words across 390 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.