@nels_ysng

Follow for Day 3! Comment “Claude” for my guide on how to use this workflow. This is Day 2 of showing you exactly how I use Claude Code to build multiple apps. #claude #claudecode #vibecoding #ai #mobileapp

You don't need to buy the AI clues course. This is day 2 of showing you exactly how I use Claude Code as a solo developer working on multiple apps. I'm not running some crazy autonomous AI agent god, but I found my own optimized workflow to use Claude Code as a coding agent while building multiple apps. First, planning. Before Claude builds anything, I give him my Cloud. MD file and relevant documentation so that it has context for what my project is and how I want my future built. I'll ask Claude to ask me clarifying questions before generating a plan. And once the plan is given back to me, I can review and approve it in my Cloud. MD file. I'll also make it a rule to ask Claude to write unit and integration tests for every major feature it implements. After it finishes implementing the feature, I then manually check and test on my local dev environment. And if there's anything I want to change or an error that I'm seeing, I paste that error into Claude code, ask it to investigate the root cause and make the fix. If you want my guide on how to use this workflow, comment Claude and I'll send it to you. Follow for day three.

Follow for Day 4! Comment “Orb” and I’ll send you the link to it. This is Day 3 of sharing tools I’m using to elevate my app’s UI/UX. #ui #vibecoding #uidesign #coding #fyp

If you don't want your app to look bad coded, I just found this free website where you can find unique and dynamic versions of common boring UI components like search bars and loading wheels. This is their thinking orb, which you can use as your loading state wheel. And the best part is you can customize the speed, size, shape, and the overall design of every component. They also have agent mascots you can use in your app's AI chatbot directly. It's compatible with React, React Native and even Swift UI, meaning you can use these components in a web, iOS and Android apps. Comment orb and I'll show you the link and how to set it up for yourself. This is day 3 of sharing tools I'm using to elevate my app's UI UX. Follow for day 4.

Follow for Day 3! Comment “UI” and I’ll send you the link to Drawably! This is Day 2 of sharing tools I’m using to elevate my app’s UI/UX. #ui #mobileapp #vibecoding #uidesign #fyp

I just found this free UI component library that can make your vibe coded app look a little bit less vibe coded. It's called drawably and it's a free and open source library of hand drawn versions of common components like buttons, toggles, checkboxes, arrows, and more. All you need to do is copy the code snippet and paste it right into your code. Or ask your coding agent to implement it for you. And what I like about it as well, they're not static components. You and generate random variants of every component. Comment UI and I'll send you the link to it and how to install it for yourself. This is day 2 of sharing tools I'm using to elevate my app's UI UX. Follow for day 3.

Follow for Day 2! Comment “Claude” for my guide on how to setup up your CLAUDE.md and project like this to manage your token usage. This is Day 1 of showing exactly how I use Claude Code to develop multiple apps. Not like the AI gurus. #vibecoding #claude #claudecode #ai #fyp

You don't need 1 million plugins to manage your Claude token usage, and you definitely don't need to buy the AI Google Script. This is day one of showing you exactly how I use Claude Code. As a solo developer working on multiple apps, every time I start a new code code session, I don't want to start by dumping a massive amount of code into the context just to develop one feature. So I create and maintain a Lean Code MD file, which basically acts as a map for my entire project. And I keep it extremely intentional. It tells code how the code base is structured, where to find things, standard patterns or conventions, any rules or any nuances it needs to know about. I don't try to cram every piece of information into this one file. Instead, I have an Agent Docs directory, and within that I have folders for every part of my app, and in those folders I have markdown files documenting specific topics that are related to that folder. I also keep an index file in each folder so that Claud knows which document to read for which topic. I explicitly tell Claude to always read documentation first and never read my code base unless absolutely necessary. And I also ask it to update existing documentation or create a new one if it doesn't exist after it implements a new feature. All of these combined keep my context much more focused, make my sessions Cheaper and gives Claude a better understanding of how I wanted to build things. Comment Claude, if you want my guide on how to structure your project like that, follow for Day 2.

Follow for Day 2! Comment “App” for my guide on how to design UI with AI. This is Day 1 of sharing tools I’m using to elevate my UI/UX. #buildinpublic #vibecoding #mobileapp #ui #fyp

I'm working on the UI of the onboarding flow for my second mobile app as a vibe coder. This is what I'm using to design it. I've been using Mobin, which is a large collection of real web and mobile app screens that you can tell Claud or Codex to look at directly with this MCP. So instead of your agent coming up with a completely random UI, it has examples of things like navigation cards, onboarding wall screens, and so much more. But the part I find even more interesting is for research. I can tell Cloud Code to do a deep dive research onto the onboarding flows of high converting apps in my category and give me a report on what they're doing right about the onboarding flows so that I can implement the same in my apps. If you want my guide on designing apps with AI, comment app and I'll send it to you. This is day 1 of sharing tools I'm using to elevate my apps UI UX follow for day 2.

Comment where you would add your rating prompt! My app’s lack of ratings may be costing half my downloads. This is what I’m doing to increase my app’s rating count. I’m adding a rating prompt after my onboarding, and only shows on the user’s second visit to the app. To force a rating, I’m not making this rating pop-up closable! Don’t worry, I’m going to let users dismiss the real Apple pop-up modal. #buildinpublic #mobileapp #vibecoding #solofounder #fyp

The lack of ratings on my app on the App Store is costing me half my downloads. My app, Jane, currently only has three ratings on the App Store because when I first launched my app, I didn't prompt my users to rate it. So I needed to change that. Apple only lets you prompt each user for a rating three times a year. So I needed to be strategic about it. Most apps will prompt for a rating as soon as you open the app or early on in the onboarding flow before the user even gets to use the app. So I added my rating prompt after my improved onboarding flow. And it's only shown if the user has opened the app more than once. And in order to force a rating from the user, I've made the rating modal not closable until the apple native rating prompt shows up where they can either give my app a rating or dismiss it completely. Let me know in the comments where you would add your rating prompt in your app. If you want to see what I improve next on my app, make sure to follow along because I'll be making more videos like this.

Comment “App” for the guide on how to launch your app on the App Store. These are 10 things you ABSOLUTELY need to do before submitting for an app review. 1. Don’t just build a web app for mobile 2. Disclose AI features/content 3. Create privacy policy and terms of use 4. Add a Delete Account button 5. Add a Restore Purchases button 6. Add Sign in With Apple 7. Don’t fake ratings or reviews in onboarding 8. Data consent policy 9. Create a website for your app 10. Create support page on website and add link to it in your app #buildinpublic #mobileapp #appstore #vibecoding #fyp

Apple has rejected close to 2 million apps in the last year. Here are 10 things you absolutely need to do before submitting your app review. No. 1 don't just build a web app for mobile. Apple is rejecting any clearly web coded apps, and honestly for good reason. Make sure you're using React, Native, Flutter, or Swift UI. No. 2 disclose any AI features or AI generated content. No. 3 create a privacy policy and terms of use with tools like Termly. No.4 add a delete button on your settings page where users can delete their account. No. 5 add a restore purchases button on your paywall so that users can restore previously purchased subscriptions or in app purchases. No. 6 if your app offers sign in, you must include a sign in with apple option. No. 7 don't fake reviews or ratings on your onboarding flow. No. 8 if you're collecting data from users, make sure you have a data consent policy. No.9 create a website for your app. A simple landing page will do. And deploy on Vercel. No. 10 create a support page on this website and add a link to it on your app setting page. Certain users are able to go for it if they require support for your app. If you want a full guide on how to launch your app on the App Store, comment app and I'll send it to you.

I started this account documenting my journey building my first mobile app, on the way to build a second but making content has also been so fun and something I’ve always wanted to do! This is how much I made in September! It’s not much but I still can’t believe I’ve earned internet money!! #fyp #contentcreator #creatoreconomy #buildinpublic

This is how much I made from content creation as a small creator in the month of September. Just saying that sentence out loud is actually crazy. I started this account documenting my journey building my first mobile app, and now I'm starting to work on my second while having a 9 to 5. And now I'm creating videos, which I've always loved ever since I was younger. And now I'm able to fund my projects and my apps with content money, which is amazing. So from my digital product, which is a guide, I earned £40 from two sales, and from brand deals, I managed to bring in £423. So that's a total of £463, which is nothing crazy, but if you told me five months ago that I was able to make even £100 creating content, I would have literally laughed in your face. I'm gonna keep sharing more about my content journey alongside the other videos that I make on building apps, tech and AI and everything else. And I'm hoping next month we can reach £1,000. So follow along if you wanna see where this goes.

Comment what you would change about my paywall and onboarding (previous update). Follow to see how I improve my app’s rating next. Last week, 13 people viewed my paywall but 0 converted into paying users. This is how I improved it: I changed it from a single screen paywall to a 2-screen paywall and increased the length of my free trial from 3 days to 1 week. I also switched from just a monthly and yearly plan, to a weekly, yearly and lifetime plan. #buildinpublic #mobileapp #vibecoding #fyp #solofounder

Last week 13 people viewed my paywall, but zero converted into paying users. So clearly something isn't working on my paywall. Multi step paywalls convert 37% better than single screen paywalls, which my app is currently using. So I split it into two steps where I show them their personalized plan and the value of Jane first and then introducing the pricing plans. I've also switched from the monthly and yearly plans to a weekly, yearly, and lifetime plan. I wanted to make the yearly plan a bit more compelling and the weekly option gives people a lower commitment way to try Jain, while the lifetime option is there for people who really want to commit. I've also increased my free trial from 3 days to 1 week. As research shows longer trials convert better than short ones. Let me know what you would change about my onboarding and paywall and follow how to see I improve my app's ratings next.

Comment “App” if you want my guide on how I launched my first mobile app. This is the exact AI coding workflow I’m going to be using to develop my second mobile app. This time, it’s different because I’ve shipped something already and I won’t make the same mistakes! First, I’m validating my idea with Peekly, seeing if there are competitors doing more than +10K MRR. I’m using ChatGPT and Claude for ideation and using Claude Code as my main coding agent. I’m using Mobbin for UI/UX research. For my frontend, I’m using SwiftUI for now, as I want the app to feel as native to iOS as possible. For the database, I’m using Supabase and for analytics, I’m using Posthog. My goal is to hit 10K a month in revenue. Let’s see where this goes shall we? #buildinpublic #vibecoding #mobileapp #coding #fyp

This is the exact AI coding workflow that I'm using to build my second app until I reach 10k a month. After 3 months of releasing my first app, I'm now working on my second, but it's different this time because I've already built and shipped something so I know what I did wrong the first time and I'm definitely not making the same mistakes again. First, I'm validating my idea with Peekly. I can use it to analyze competitor and winning apps and look into their monthly recurring revenue and check their keyword rankings for app store optimization. I haven't tried this yet, but you can also look at the meta ads that are running for different apps. If I find multiple similar apps doing more than 10k a month in MRR, that's a pretty good signal that there's demand and my idea is validated. For my front end, I'm gonna be switching from React Native to swiftui because I want this app to feel as native to apple as possible. I'm using ChatGPT and Claude for ideation and using Claude Design to explore and design my UI. For my main coding agent, I'm using Cloud Code. I'm also using Mobbin and its MCP on cloud code to research existing apps and pull in UI references. For my back end, I'll be using Superbase and for analytics, I'll be sticking to Posthog as it's worked well for my first app. So this is officially app No. 2 and the Goal is to build and ship it and eventually reach 10K a month in revenue. If you want a guide on how I launched my first mobile app on the App Store, comment app and I'll send it to you. If you're building something too, make sure to follow along because you'll be making more videos like this as I built my second app in public.

Comment “App” for my guide on how to launch your first mobile app. Follow to see how I improve my paywall next! Last week, only 34% of people who started my onboarding actually finished it. That’s when I knew something was wrong with it! I used Mobbin MCP to research high converting onboarding flows and the biggest change I made was including interactive feature usage in my onboarding flow. That way, users can get to use some of the features of the app before seeing a paywall. #buildinpublic #coding #mobileapp #vibecoding #fyp

I released my first mobile app three months ago and we're almost at 100 downloads and 3 paying users. But looking at my analytics on Posthog, I realized that my onboarding kind of sucks. Over the last week, only 34% of the people who started my onboarding actually finished it. So clearly something isn't working. So I use Mobbin MCP on Cloud Code to research high converting onboarding flows in apps in my category. And the biggest change that I made to my onboarding was adding interactive steps where users actually get to use some of the app's features. So instead of just telling users what my app Jane does, they can create their first speaking situation and log their confidence. And the idea is that they can actually experience the app during onboarding and hopefully get their first win moment before even seeing the paywall. After their first win moment, I then ask for notification permission, which will hopefully help with my app's retention. Next is to enhance my paywall, so follow to see how I improve that. If you want my guide on how to release your first mobile app, comment app and I'll send it to you.

Use the link below or in my bio for 200 free credits to start building on Enter Pro! https://enter.converge.ai/s/qW22s2 I've used many AI app builders but Enter Pro is genuinely the only one that has ticked all of my boxes as a solo app developer. @Enter Pro by Converge.AI #EnterPro #convergeai #ai #vibecoding #fyp

I've tried a lot of AI app builders, but this is the first one that's actually takes all of my boxes as a solo app developer. It's called Enter Pro, and I wanted to put it to the test by building a landing page from scratch. I decided to build a landing page for a digital museum, a space where I can showcase the different ideas and projects that I'm constantly coming up with as a solo app builder. So I started by giving it a prompt and letting it understand what I wanted built. Even better, you can start off by browsing a ton of templates and using that as a starting point. And what I really like is that it doesn't start generating everything immediately. With plan mode, it maps out the build first and I can review the approach before it starts building anything. And because I'm pretty particular about design, being able to edit the app directly through the visual editor is very useful. I can make changes to the layout and interfaces in real time and even use voice input when I don't want to type or touch the code. I also used the skill system and gave the AI additional capabilities and made it better suited to the type of project that I was building. And after a few iterations, this is what I ended up with. A fully working interactive landing page that I can deploy and share in one click. If you Want to try enter Pro yourself? Use my link below for 200 free credits to start building!

Comment “Claude” for the complete guide on the 6 Levels To Using Claude Code. Level 5. Parallel Agent Execution: In this level, you’re running multiple Claude subagents on different Git branches working on several different features for example. You can use tools like Warp or Superset to help run and manage subagents. Level 6. Connecting With Outside World: In this final level, you’re using MCPs to connect to external frameworks, apps or softwares. For example, you can use Supabase MCP for Claude Code to directly read data in your database instead of giving it your schema and data manually. With MCPs, you can start taking your development with Claude Code to the next level. That’s the full journey! Let me know if I’ve missed anything in these 3 parts. #ai #claudecode #vibecoding #claude #fyp

Welcome to the final part of my series, 6 levels to using cloud code. Level 5 is where most people that are watching haven't discovered yet sub agents and parallel execution. And here's where you're running multiple cloud code agents at one time, each working on a different feature, fixing bugs or working on testing. Each agent works on his own Git branch in a Git work tree. If you combine that with the skills that you've created, sub agents are pulling in specific skills to execute specific tasks. So how do you run parallel execution? Use tools like warp or superset to run and manage your subagents. The trap here is that many people skip the first 4 levels and start running agents parallel without the proper foundations. So no lean cloud MD file, no custom skills or workflows to allow subagents to work optimally. Level 6, connecting with the outside world. Level 6 is where you start using MCPS to connect to external frameworks and applications. For example, using Posthog MCP for Claude code to look into your analytics directly instead of just guessing. So once you've reached level 6, that's the entire journey. Claude Code went from just another chatbot to being part of your entire development journey. So what level are you on right now? And let me know if I've missed anything in this series. If you want a full guide on the six levels of using Claude code, comment Claude and I'll send that To you.

Comment “CLAUDE” for the full guide on how to write your own CLAUDE.md file. Follow for part 3!!! The trap we covered in Part 1 in Levels 1 and 2 was the lack of context. Going from Level 2 to 3 changes that. Level 3. Full Context: You’re on this level if you write and maintain your own CLAUDE.md file. The CLAUDE.md is the first file Claude Code reads when you open a new session. Here, you would have your project’s architecture, standard conventions, rules and more. Now, Claude Code isn’t guessing anymore and it doesn’t have to read your entire codebase to understand your project. You’re also on Level 3 if you use /clear to clear your context window. The more you chat in one session, the worse Claude Code works because the context window is limited. Claude Code has a /compact command, however, I don’t use it because it uses tokens to summarise your conversation and brings the context to a new session, so I rather use /clear. Level 4. Skills and Workflows: A Claude Skill is a folder with a specific playbook, like you can have a specific skill for handling authentication in your app. You can use Claude’s Skill Creator skill to create your first skill. On top of skills, you can build workflows. Create custom slash commands for repeat tasks. Between workflow and skills, you’re not just prompting anymore, you’re designing the pipeline it follows every time. Follow for levels 5 and 6, where you leverage subagents and parallel execution. #vibecoding #claude #claudecode #ai #fyp

This is part 2 of my series 6 levels to using Claude code. The jump from level 2 to level 3 will change how you work with Claude code. Level 3 full context you're here if you write and maintain your own Claude MD files. The Claude MD file is the first file that Claude ever reads when you open a new session. In the Claude file you'll have your products, architecture, tech stack, standard conventions, standard patterns, what to do, what not to do. Now Claude isn't guessing your patterns, it's following them. In this level you're also using slash clear to manage your context window, because the more you chat with Claude in one session, the worse his output becomes. Now Claude does have a slash compact command which summarizes your session and opens the new chat with that context. But I don't like doing this because it just eats up our tokens. So honestly it's just better to use slash clear. The trap in this level is a bloated Claude MD file. It's much better to have a lean few shot example Claude MD file rather than long instructions. So instead of using slash init for Claude to create a Claude MD file, I suggest writing yourself because less is more. If you want a guide on how to write your own Claude MD file, comment Claude and I'll send it to you. Level 4 skills and workflows. This is where you start using skills to encode Your standards and workflows. So instead of having one cloud endpoint for rules and standards for everything, A skill is essentially a folder with a specific playbook. Like you can have a specific skill for a back end authentication. For example, on top of skills, you can create workflows and create custom slash commands for repeated tasks. Between workflows and skills, you're not just prompting anymore. You're designing the process. It follows every single time. But the trap is having too many skills. So it's better to have 20 to 30 curated skills rather than 100 generic skills. Once again, less is more. Make sure you follow for levels 5 and 6, where you're gonna be leveraging sub agents and parallel execution.

Follow for part 2 tomorrow!! There are 6 levels to using Claude Code. Most people are stuck on these first 2 levels and don’t even know it. Level 1. Chatbot Crutch: You’re on this level if you treat Claude Code like another chatbot. You see Claude as a Command -> Output. No context or setup. This is the trap. Level 2. Partner: On this level, you’re using Plan mode and working with Claude Code as a partner and letting it push back on your ideas. However, the trap is still the same. Claude is still guessing about your project, stack and preferences. That’s where we go to Level 3 and 4 which I’ll be getting in to tomorrow. The jump from level 2 to level 3 will change how you fundamentally use Claude Code. #vibecoding #claude #claudecode #fyp #ai

There are 6 levels to using Claude code and most people are stuck on the first two levels and don't even know it. Level 1, the chatbot crutch. And this level treats Claude code like another chatbot. Open Claude code and immediately start prompting. No setup, no context, and all you see Claude as is just command output. The trap is that there's no context, so Claude is just guessing everything. And guessing means Claude is using the average of the training data leading to slop. So that's where we go to level 2. Level 2, the partner. You're on level 2 if you're using plan mode before building anything and you discuss with Claude code before touching a single line of code. And the skills you master here are plan mode, extended thinking, working with Claude Code as a partner and letting Claude push back on your ideas. But the trap is still the same. Claude is still guessing about your project, your stack, and your preferences. Words aren't enough. This is part 1 of my series, 6 levels of using Claude Code. Make sure you follow to see how you go from level 2 to levels 3 and four.

Comment “Automation” if you want the GitHub link to set this up for yourself!! I built a Manychat clone because as a small creator, I didn’t want to pay for it. #fyp #buildinpublic #opensource #vibecoding #contentcreator

I didn't want to pay for manychat, so I built my own and I'm making it free and open source for anyone to use now. Don't get me wrong, manychat is great, but to be honest, I only needed one feature which was the commented DM automation, and the pricing plans just didn't look worth it to me. And you've probably seen many developers promoting manychat. I'm calling them out because they know that you can do it for free. So I started working with Cloud Code to help me build a manychat clone. No fluff, just a core functionality to connect my Instagram and create comment to DM automations. this almost failed because meta for developers was a pain to set up as I wasn't able to receive SMS notifications to verify my Facebook account to create a developer account. However, after reading countless meta community posts and just waiting for one day, it finally worked. So I was so happy I went back to going to work on it. But on day 2 I faced another problem. In order to set up and use Instagram's APIs, you need to create and publish a meta app. And in order to publish the app, you need to add a privacy policy URL. Which you aren't able to because the basic setting screen that meta tells you to add it into is not loading. It's. It's. It's just not loading at all. And this has been an ongoing bug for A few months now. So meta. Please fix it. I was so annoyed and I almost gave up, but the thought of paying for a subscription for something that I can do myself infuriated me. So I kept digging and went through the dashboard and I found a loophole. I was able to add the privacy policy URL into the AT review section of the dashboard. So once I was able to figure that out, I was able to call the API and the automation worked. And after testing locally I deployed it on versel and it's now running perfectly and I've been using it for the past few weeks. So if you want the Gethub link and set it up for yourself, comment automation and I'll send it to you.

5 months ago, I posted my first video. I opened up to the internet about my stutter and released an app that helps people who stutter build their confidence and started creating content about it. Today, that app has +100 downloads, and several paying users and a 1000 of you decided to follow my journey and learn how to build apps with me. I’ve made some very cool connections on here, other developers, builders and people on a similar journey and have been able to start collaborating with brands as well!! If you told me this was possible 5 months ago, I would’ve literally laughed in your face. I’m so glad I started posting content. If you’re on the fence about posting content, trust me, just do it. The hardest part was starting :) #buildinpublic #mobileapp #solofounder #coding #fyp

Comment “App” for the full detailed guide on how to setup subscriptions for your mobile app with RevenueCat!! Follow for more tips and tutorials on launching your first app! 1. On App Store Connect, create a subscription group and add your products to that subscription group. Give each a unique product ID and keep the naming convention consistent across different products. 2. Set your pricing, localisations for display name and submit it for review. 3. Go onto RevenueCat and connect your app to RevenueCat using the App Store Connect API key. RevenueCat will then pull in your products automatically. 4. Create an entitlement 5. Link an offering to the entitlement. Typically, you can just use the default offering. 6. Lastly, test your subscriptions with StoreKit by adding a StoreKit configuration file on Xcode and mirror the product IDs and run your tests on a sandbox tester account. #fyp #buildinpublic #mobileapp #coding #applaunch

Here's how to set up subscriptions for your iOS app in under 60 seconds. First, in App Store connect, create your subscription group, then add your products. Give each a unique product ID, as this ID is what will connect everything later, so keep it consistent. So I suggest something like this. And then set your pricing, localize a display name, then submit it for review. Second, we're gonna use Revenue Cat, so if you haven't already, create an account, add your app onto Revenue Cat using the App Store Connect API key, which you can find here. Revenue Cat will then pull your product IDS automatically. Next, create an entitlement, and this is the thing that the user unlocks, for example, premium. After that, create an offering, which is the paywall configuration you will show to your users and attach to your products as packages. And lastly, test with Storekit. Add a Storekit configuration file on Xcode, mirror your product IDS and run on a sandbox tester account on a real device. On Xcode, use the debug tab to test your store kit subscriptions and test what happens when the user renews or cancels your subscription. If you want a full guide on how to set up subscriptions on your app, comment app and I'll send it to you.

Comment “App” for a full detailed guide on how to release a mobile app on the App Store. I just hit 3 paying users for my first mobile app. Here are 3 things I would do differently if I could start over. I’m taking these learnings with me as I continue my mobile app development journey. #fyp #buildinpublic #mobileapp #vibecoding #solofounder

I just hit three paying users for my first mobile app that I released two months ago, and here are three things I would do differently if I had to start over. Firstly, I would validate my idea with real people and not just doing research before building anything. I did validate my idea, sort of. I did a deep dive on Reddit figuring out people who stutter actually wanted an app to help with their stuttering. And I also did research on every other stuttering app that there was, looking at the pros and the cons. But what I skipped was talking to real potential users directly. And now I'm playing catch up, trying to market my way to users after the fact. So if I started over, I would spend some time gathering a waitlist of real users. Because research tells you an idea could work, but a waitlist will tell you if people will actually use your app. No. 2 implement analytics from day one. After a month of release, I hit a wall. I didn't know what to build next or fix or improve. And that's when I finally integrated Posthog Analytics into my application to analyze user journeys, engagement, retention and errors and all that kind of stuff. But by then it was already too late. I had downloads, but those users had already stopped using the app and I had no idea why. No. 3 run a close beta testing before launching your app. Get people who aren't you to use your app so it could be friends or family, and that way you can catch bugs before they go into production. I had to find this out the hard way. So if you're building an iOS app, you can use Testflight and invite people as testers on App Store Connect. If you want a full guide on how to release an app on the App Store, Comment app and I'll send it to you. I'm Nelson and I'm documenting my journey building my first mobile app. I'll be making more of these kinds of videos, so if you're on a similar journey, make sure you follow along.

Make sure to follow for part 3! Comment “App” if you want a full guide on how to launch an app on the App Store. Here are 6 more things you need to do before submitting your app for an App Store review. #mobileapp #fyp #buildinpublic #applaunch #coding

If you're launching on the App Store for the first time, this video is for you. My app took one month to get approved by Apple's app review, so here's part 2 of 6 things you need to do before submitting for an app review so yours won't take so long. No. 1 if your app offers in app purchases or subscriptions, you need to add a restore purchases and it's very important on your paywall screen. And this button will be used by users who already have a subscription to restore their purchases and get back into the app. No. 2 make sure you fill out details on App Store connect accurately, double check, triple check, and don't try to be sneaky here. This goes for data privacy, app privacy, and age ratings. Inaccurate data could lead to multiple rejections or even worse, your app being taken down completely. No. 3 leave detailed notes for the reviewer in the note section of the App Review. Every day apple gets a ton of submissions for app review, so explain anything that might not be obvious or any quirks. Apple's reviewers move fast and reject on any ambiguity. So add a few sentences because even that can go a long way. No.4 don't gate core features behind account creation if you don't need to. Apple may push back on unnecessary friction and also is not great for the user's experience anyway. I Learned this mistake by making signing mandatory and Seeing many users drop off on a signing screen. No. 5 make sure your subscriptions and in app purchases are configured correctly. Make sure to test them on store kit on a real device. Make sure they're in a ready to submit state before you submit for an app review. No. 6 if your app has any AI features or AI generated content, you must disclose it clearly and be upfront about it on the metadata and on the app's UI. This will avoid claims that could be misleading about accuracy or functionality. If you want a full guide on how to launch an app on the App Store, comment app and I'll send that to you. And make sure you follow for part 3.