Bootstrapping an AI App to $2M ARR

Summary

Podcast Summary: Bootstrapping an AI App to $2M ARR

Podcast Information:

• Title: App Marketing by App Masters
• Host: Steve P. Young
• Episode: Bootstrapping an AI App to $2M ARR
• Guest: Josh Moore

---

1. Introduction

In this episode of App Marketing by App Masters, host Steve P. Young welcomes Josh Moore, an indie app developer who successfully built and scaled his AI-powered app to $2 million in Annual Recurring Revenue (ARR) within the first year. The conversation delves into Josh's journey, strategies, and insights that led to his rapid growth.

[00:11] Josh Moore: "Are you looking to make your in-app messaging more unique? Well, you can elevate your communication with Foggy custom emojis..."

(Note: Advertisement segments are omitted from this summary.)

2. Origin Story: From Idea to Reality

Josh Moore shares his motivation and the initial steps taken to develop his AI app, Wave AI.

[02:16] Josh Moore: "With ChatGPT, when it came out about a year and a half ago, was like that bridge that I really needed..."

Josh explains how observing his father's challenges during medical appointments inspired him to create an app that records, transcribes, and summarizes conversations. This real-world need highlighted the potential of AI in simplifying complex tasks.

3. Leveraging AI for App Development

Josh emphasizes the pivotal role of AI in both the creation and functionality of Wave AI.

[05:04] Josh Moore: "Did you use AI to completely build this app too? Is that what I'm hearing?"

[05:09] Josh Moore: "That is what you're hearing, yeah. AI was super part of the origin story."

As a non-engineer, Josh utilized AI tools like ChatGPT to bridge his knowledge gaps, enabling him to independently develop the app without a formal engineering background.

4. Monetization Strategies

From the onset, Josh integrated a monetization model to ensure sustainability.

[06:26] Josh Moore: "I really didn't think anyone would buy it... But part of the project was getting something live in the App Store, which to me included having an option for anyone to download it and buy it if they wanted."

Josh implemented a freemium model, offering free accounts with limited usage to attract users while encouraging paid subscriptions for enhanced features. This approach proved effective as user feedback began to pour in, validating the app's market fit.

5. Scaling Through Strategic Marketing

Josh outlines his marketing tactics that contributed significantly to Wave AI’s growth.

[09:05] Josh Moore: "I started with Apple search ads. That's sort of bottom of the funnel... then meta after that..."

He utilized Apple Search Ads to target users actively searching for related keywords, followed by Meta (Facebook) Ads to broaden his reach. By focusing on blended acquisition rates, Josh ensured that his ad spend was sustainable and profitable from day one.

6. User Acquisition and Advertising

The discussion delves deeper into Josh’s advertising strategies and the importance of managing ad spend effectively.

[10:47] Josh Moore: "Meta has been really kind of amazing in that we send it enough signals for it to kind of figure out who my users are and it finds them extremely efficiently."

Josh emphasizes the significance of using the Facebook SDK over traditional Mobile Measurement Partners (MMPs) for better performance tracking. This choice allowed him to streamline his ad campaigns and optimize for high-conversion events like subscriptions.

7. Paywall and Pricing Tactics

Josh shares his insights on setting up effective paywalls and pricing models.

[17:57] Steve P. Young: "I have 30 minutes free per month..."

[18:27] Steve P. Young: "...put an annual in the middle. It was on the right until like yesterday and we did a test at the middle and it's so much better."

He experimented with different paywall placements and subscription models, finding that immediate paywalls without free trials often resulted in higher conversion rates. By offering limited free usage and emphasizing value, Josh encouraged users to commit to subscriptions promptly.

8. Technical Challenges and Solutions

Josh discusses the technical hurdles faced during app development and how he overcame them.

[42:43] Josh Moore: "Could you expand more on hosting your own AI?"

[42:49] Steve P. Young: "Whisper, the OpenAI transcription model, is open source... running Whisper on my own rented hardware would be cheaper and better."

By hosting the Whisper transcription model himself, Josh reduced costs and gained greater control over functionality, such as speaker recognition. This self-hosted approach enabled him to handle long-form recordings without the limitations imposed by third-party APIs.

9. Community Engagement and Feedback

Engaging directly with users and the broader community played a crucial role in Wave AI’s success.

[08:48] Josh Moore: "I have a 4.9 star, about 2,000 ratings. It was important to me from the get go to be good and to like have it be useful to people..."

High user ratings and active feedback through channels like Intercom provided Josh with valuable insights, allowing him to refine the app continually. This responsiveness fostered a loyal user base and enhanced the app’s reputation.

10. Q&A Session: Key Takeaways

During the Q&A segment, Josh addresses various audience questions, offering actionable advice for app marketers and developers.

a. Managing Ad Budget and ROI

[50:37] Steve P. Young: "I started with $300 a day... I manage spend by spending double or half based on results."

Josh advises maintaining a flexible ad budget that aligns with ROI goals, ensuring that scaling efforts remain financially viable without risking significant losses.

b. Organic vs. Paid Downloads

Understanding the interplay between organic and paid downloads is vital.

[65:42] Josh Moore: "If you could stop ads, look at your all downloads, because it'll give you an indication of how well Facebook is performing."

Josh recommends evaluating the overall impact of ads on both paid and organic downloads to gauge the effectiveness of marketing strategies.

c. Optimizing Token Usage

Technical efficiency in AI operations can significantly affect costs.

[47:41] Josh Moore: "How do I minimize token usage?"

[47:41] Steve P. Young: "I want to use the best APIs for the best summary... Use all the tokens you want."

Josh emphasizes prioritizing quality over minimal token usage, ensuring the app delivers superior performance and user experience.

11. Final Insights and Future Plans

As the episode concludes, Josh reflects on his journey and shares his vision for the future.

[66:43] Josh Moore: "How are you in time?"

[67:17] Steve P. Young: "Wave Co and then Wave AI in the app stores."

Josh plans to continue refining Wave AI, exploring additional marketing channels like Google Ads for Android, and enhancing app functionalities based on user feedback. His commitment to quality and user satisfaction remains central to his strategy moving forward.

Notable Quotes

• Josh Moore [02:16]: "With ChatGPT... was like that bridge that I really needed..."
• Steve P. Young [05:09]: "AI was super part of the origin story."
• Steve P. Young [17:57]: "I have 30 minutes free per month..."
• Josh Moore [10:47]: "Meta has been really kind of amazing..."
• Steve P. Young [18:27]: "...put an annual in the middle..."

Conclusion

Josh Moore’s successful trajectory with Wave AI serves as an inspiring case study for indie developers and app marketers. By leveraging AI, strategic marketing, effective monetization, and robust technical solutions, Josh achieved impressive growth without external funding. His insights offer valuable lessons on balancing user acquisition, maintaining quality, and ensuring sustainable scaling in the competitive app market.

For more insights and strategies on app marketing, tune into future episodes of App Marketing by App Masters every Friday at 9 AM Pacific on YouTube.

Transcript

[Steve P. Young]

The most action packed content from the top mobile experts. This is the App Masters podcast with Steve P. Young.

[Josh Moore]

Are you looking to make your in app messaging more unique? Well, you can elevate your communication with Foggy custom emojis. Designed from authentic content, their custom emojis convey genuine emotions and meanings. Whether it's for individual chats or mass communication, Fogy seamlessly integrates across all your favorite platforms. Learn more by going to fogiapp. Com. That once again is fogyapp. Com. What is up, app nation? It is Steve P. Young, founder of AppMasters.com and you are in the right place every Friday, 9am Pacific on YouTube. If you're looking to grow your app downloads and more importantly, those revenues. And today, speaking of revenues, this indie has X Uber as well. He's built an app and he's scaled it all the way to 2 million ARR within his first year. I want to talk about how he's been able to do it so freaking fast, what he's learned throughout the entire journey. So without further ado, let's bring in Josh Moore. Josh, welcome in. Thanks.

[Steve P. Young]

I wish that would pop up every time I joined any kind of like video conference, you know, just right.

[Josh Moore]

Let them know, Let them know you're here.

[Steve P. Young]

That's amazing. That's amazing. Thanks for having.

[Josh Moore]

Congrats on the success, man.

[Steve P. Young]

Thank you. I owe it all to you.

[Josh Moore]

I think I'll take all the credit, man. You heard it first. Hey, I'm gonna share the app, so if you guys want to check it out, it is Wave Co. And I'm gonna pull that up as well. Wave Co. And then just search for Wave or Wave AI in the App Store. It's a note taking app. It's AI generated. Josh, maybe we start off with maybe where you came up with this idea and I definitely want to talk about how you've been able to scale it so fast as well.

[Steve P. Young]

Yeah, so, you know, I've, I've always wanted to learn to code, as they say, and I've taken a few shots at it, but rolling the tape all the way back. I've always been into computers and the Internet, but came up on the business side. Started selling stuff over the Internet when I was in high school in the 90s. Eventually about a decade later, found my way to Uber where I was an early employee there leading the New York City business. And you know, throughout all those experiences, I kind of viewed engineering as potentially a more fun role when you're building a product that's used online, you know, and so I've tried a couple times before to like, make an app, try something out. And ChatGPT, when it came out about a year and a half ago, was like that bridge that I really needed, you know, something that I could ask a bunch of questions to. They could help me learn about different things. And so with that, I started building little small apps. You know, they would do different things. And one of them was, and it was a voice recorder that would transcribe the audio and then summarize that. And it was a neat little trick. You know, you got to lunch with someone and put it on the, you know, run it for a bit and then show them what to do. But the real aha moment was when I gave it to my dad, who was going to the doctor, and he was a doctor himself before he retired, but would go to these appointments and they would be full of sort of technical speak, and it goes really fast and would have trouble telling the rest of the family what was talked about. So just for fun, I said, hey, I'm working on this thing. You know, there's plenty of audio recorders for Wave meetings. I'm sorry, for Zoom meetings and meets and teams and all this stuff. But what about just out in the world? Try this out. Brings it records the visit. And it was really eye opening. It was like, wow, that's really incredible. And I found at that time, it still may be somewhat the case that a lot of the best AI tools and demos and things are happening in web browsers and at desks and in places that don't really unlock them to the whole world. And so Wave is a very simple app. It records audio, it transcribes it and summarizes that. It also works for phone, so you can make a phone call through the app, just like a regular phone call. It'll record that and summarize that. You can import podcasts and YouTube videos and some other things that I'm working on. And it's had a bit of success in the App Store. And so, I mean, I'll. I'll stop there. That was a lot. But that is. That's the origin story.

[Josh Moore]

Did you use AI to completely build this app too? Is that what I'm hearing?

[Steve P. Young]

That is what you're hearing, yeah. That. I think it's a little confusing because it's an AI app, but AI is super part of the origin story. I mean, I came up on the business side. This is my first engineering work. I built it alone. And yeah, AI was the bridge for that as well. So it's sort of AI on both sides.

[Josh Moore]

Wow, that's incredible. Okay, I love it and I love the aha moment. Did you have monetization when your dad was playing around with it too, or how?

[Steve P. Young]

Well, I gave him a. I gave him a free account, sort of as, like, reciprocity for, you know, raising me. You know, as many of your listeners and viewers probably know, AI software is a little different than regular software in that the processing is not. It has, it has. It's not free, it's come down in Price. But running ChatGPT or Transcription Engine, excuse me, is not free. So I had monetization from the beginning. I also sort of before. At this point, I didn't really think anyone would want it, but part of the project was getting something live in the App Store, which to me included having an option for anyone to download it and buy it if they wanted. I really didn't actually kind of think anyone would buy it.

[Josh Moore]

I love it.

[Steve P. Young]

So it was, it was just for fun. And then about a year ago, probably like last July or August, the feedback started coming in. That reminded me of feedback I'd seen at Uber. And not to say that this is like that at all. The real, like, magic isn't built by me, it's built mostly by OpenAI. But I'm delivering the magic and some of the feedback. Feedback. I'd get an intercom in the app because I do. All the support also had the kind of excitement vibe that I remembered at Uber and remember, I joined there before anyone had ever heard of it. And so I was there for understanding that it was going to be a very big deal. And so hearing some of that, I think really AI large language models, the best feature of them in my mind at the moment is summarization. And so for a normal who maybe isn't bathing in the AI world every day when they can record a meeting at their desk and get bullets or a good, like, summary back, that does feel like magic the first time because it's like, oh, I didn't know a computer knew how to do that. They could, like, take the audio and make words and maybe they'll be right or maybe not. But to actually synthesize summarization is kind of a new thing that computers can do well. And so I felt like bringing that to normies was the way to go.

[Josh Moore]

Interesting. And you were getting this. So you had intercom within the first version of the app and you were getting this type of feedback. It wasn't even like they were just like, wow, this is impressive, Josh. Like, they got. You got that type of feedback on the. It was.

[Steve P. Young]

There was some of that. There was some. Like, sometimes you can tell the intensity of someone's anger when the app works wrong or it doesn't work or it stops working. Like, hate and love are neighbors. They're not opposites. And so, you know, I would get things like, like, I'm so mad you lost my recording. And I'd be like, oh, it's fixed. And be like, oh, thank God. I love this app. It's the best app in the world. I'm like, okay, got it. And certainly it's not for everybody and people churn and all this stuff, but I have gotten a sense that it's a basic and unopinionated tool. So it doesn't insist on being embedded in a certain way or done on a computer. It just sort of like hit, hit play and just go. And that seems to work.

[Josh Moore]

Okay, talk to me about the early days. I'm always curious because I never know. I guess the general question, high level, without getting too specific, is like, when do you know is the right time to scale? Like, when do you start spending money on ads? Do you do it right from the jump? Like, what was your thought process?

[Steve P. Young]

Yeah, so my background is weird. Like, I think you probably have a lot of instances out in the world where someone is a veteran engineer and they dabble in the, like, marketing stuff. That's probably common. Ish. I've sort of done the flip side, so I was comfortable with acquisition stuff right away. But there's a lot of nuance in the iOS world. Much of it I learned about from you, from, like, recordings and, you know, the old days was like last week kind of. But this is really a few months ago, learning about iOS 14.5 and how you can't really track and all this stuff. But my thinking was if I could have, you know, I don't want to go into the red on ads. If I can unbalance, get to a blended acquisition rate that gets paid back at activation, you know, so I'm making the money back on average on day zero or on month zero, I guess then it's worth scaling until that isn't true anymore. And so I started with Apple search ads. That's sort of bottom of the funnel. Easiest way just to get someone to see your ad. I mean, to see your app, if they search for anything kind of in your world, highly relevant, the person is ready to go. And then meta after that, which I got some help on. And that's a whole new world. You know, the whole user generated video angle which I think dominates and the hook and getting, I mean really just doing video ads is frankly new for me. When I was doing ads that didn't really look that way. But meta has been really kind of amazing in that we send it enough signals for it to kind of figure out who my users are and it finds them extremely efficiently.

[Josh Moore]

That's interesting. And Josh, did you look at. So were you kind of like in the early days, did you have. This is a common question I get from people all the time. Did you have a budget in mind? How much did you want to spend to make sure that you were had enough volume to test your hypothesis of that? What'd you call it? Day one, day zero activation payback rate?

[Steve P. Young]

Yeah, I got the sense from listening to you and some others that yeah, like if you couldn't start meta at a big enough volume, it wouldn't really have a chance to learn. And so I think we were pro. I was probably doing like 10k mrr before I started doing ads. That was when it was like, okay, maybe this would work. And then I just, I never wanted to have a month where I went deep red. Like I can afford for it to be break even. And it was in those early months. But I think I've been clear for myself about what the goal is of this. Like I don't have any VC money and I don't plan to change that. I don't really want to have a team. I have a couple folks working for the app but I think like it's not going to have a large employee base. And so, you know, the goal is to make money. So I don't want to burn that all on ads. So I'm trying to be, you know, careful about how, how big I go and try to keep the results in focus. You know, it's, it's, it's, it's easy to go crazy on that stuff, if that makes sense.

[Josh Moore]

How did you get it to that? Now I get to. How did you get it? Just organically.

[Steve P. Young]

How'd I get tick R? Yeah, I mean on Twitter I started talking about it. Actually LinkedIn was a channel where I got kind of way more react actions and I think I started putting graphs up. I mean I think that might be how we set this up. I feel like it's a cheap way to get some attention to just put your data out there. I don't think there's a person out there who's going to be turned on or inspired to come and do the same thing as me because they see the graph that I have. So I think it's pretty low risk and it gets some attention. And so people were curious. I mean, I think maybe less about what the app does and more about an old business veteran with no engineering experience making an app like either that's like a terrible idea and I want to kind of see the bloodbath, like, or maybe it's interesting. And so perhaps from those two groups people started to download it. But I think like the way is to be good. You know, I'm a 4.9 star, about 2,000 ratings. Like it was important to me from the get go to be good and to like have it be useful to people and not lose their audio. And that's like I still aspire in that direction. Like I wanted to work all the time. It still has some issues. I want it to be as useful as it can to as many folks as possible. But I do kind of think like after a certain point, just being good is the answer.

[Josh Moore]

So if I had. No, go ahead. Sorry, Josh, I didn't mean to interrupt.

[Steve P. Young]

No, I see a competitor who's like a 4.6 and this is just kind of like made up. I know there's probably data on this that doesn't matter, but I think it's just reflective of are people having a good time in that app? Are they getting done what they want? Is it working the way they expect? And the rating reflects that. And so I think, I think that's part of it. I think just sort of being high quality is that really matters. I think that's sort of like the promise of the Internet in a lot of ways.

[Josh Moore]

I love it.

[Steve P. Young]

Good things find their way to the top.

[Josh Moore]

So if I sort of, if I were to try to replicate Josh's success, it would be build an app like an mvp. Good. To build a good app, put some intercom messaging in there, look for the signals, right? Like the excitement, the love, the hate. And then yeah, your success product, market fit.

[Steve P. Young]

I also think I generally, this is sort of my first zero to one thing. Like I've done a lot of one to ends, but I haven't done any zero to ones. And I think it was a little bit easier than I imagined. Like I really thought it was. I thought I was gonna able to do it. I was like this is not gonna work. But I think people out there are surprisingly willing to try things, spend a few bucks, give it a shot, and are tolerant of imperfection. And it was just, you know, the 0 to 1 was easier than I thought. So I think, like, just getting something out there. I had good friends who are like, this is not what you're building is not useful. Like, it's not a credible. It's not enough. It's like. And maybe there's some truth to that long term, especially as some of these features end up in iOS 18. But I kind of think I'm going to do it better. And for the person who needs it and depends on it, they're going to be willing to spend a few bucks to get it. So it's just finding those people. Like, the Internet is very good at helping you find the people out there. Let's say I need, you know, I have, like, about 10,000 subscribers now. And, like, I think if I had like, 100,000, that would be a good day's work. Like, I'd be good to go with that. A hundred thousand people out of 300 million Americans and 8 billion humans is actually, like, not that crazy. It's a lot, but it's not that crazy. It is doable. So that's what gets me going.

[Josh Moore]

I love it. Do you think luck played into this at all? Because you said zero to one was so easy.

[Steve P. Young]

You know, I spent some time as a VC and I've heard a lot of ideas, and I think many ideas are very bad, and I still hear plenty of bad ideas. And I guess maybe, like, the bar for me to want to go do something was high, but I put out some ideas, too. Like, I had bad ideas, too. This wasn't really the first thing. There's a couple. So, you know, and certainly luck, like, I'm high profile because of other stuff, and so I think that's probably helpful, too. But again, like, that was certainly. That made it easier. And luck is always part of success, I think. But it's. I think it really could be anybody.

[Josh Moore]

Yeah. Honestly, I was just curious what you would say. I didn't think it was luck. I'm Josh. Okay.

[Steve P. Young]

It's a little bit of luck. Of course it's a little bit of luck. Everything is a little bit luck.

[Josh Moore]

A little bit.

[Steve P. Young]

The Uber thing was luck that I was born when I was at this time, in this moment. And it's all luck, man. Yeah, a lot of it is luck.

[Josh Moore]

I won't say. Hi, Rose here. Good to see you. Neurix ideas. Steve. How are you, bro? Good to see you, bro. Crypto crunch. Louise, Patrick is here. Sebastian. All right, good. Emory. Hey, Steve. So crowded today. Yeah, that's good.

[Steve P. Young]

Uh.

[Josh Moore]

Oh, I don't know what that means. Vaklay says, hey, Steve, I have an application whose business model is subscription based. Now I'm trying a hard paywall. Should the subscription have a trial period when yes and when not. What do you think, Josh? Do you have a trial? You have a trial, right? Or no, I don't.

[Steve P. Young]

I have 30 minutes free per month. But you know, there have been so many things I've learned from you and others that are. That have not been intuitive. And I'll tell you some of them. One is like, make sure everyone sees the paywall. Like shove it in their face. Right. Be an annual plan. Those are two things that I find very uncomfortable but that have yielded amazing success. So, you know, it's kind of crazy. Yeah, you're seeing the paywall. Definitely. Let me track you. Let me say push notifications. Okay. So yeah, everyone is seeing this until they pay. You can exit and still use the app up to 30 minutes a month, but everyone's seeing this. Oh, shit. You don't have to do that.

[Josh Moore]

No, I was just wondering if you had a trial. That's all I was picking out.

[Steve P. Young]

Oh, I know, yeah. Now, little UI bug. Excellent. So phone call recording, you're going through all of it. But yeah, there have definitely been things. A trial is probably. I think that that works. Marcus Burke, who does some work with me on user acquisition, taugh me that trials are great. One downside is that, and I might get this wrong, I know he was a guest of yours before and probably said this, but the signals you get from a trial are definitionally lower value than a subscription than someone who actually buys.

[Josh Moore]

Okay.

[Steve P. Young]

So by forcing the subscription to happen or not right up front, we're potentially giving meta better signal.

[Josh Moore]

Interesting.

[Steve P. Young]

And so that's sort of his. One of his thoughts about it. We put annual in the middle. It was on the right until like yesterday and we did a test at the middle and it's so much better. These things make no sense, but, you know, it is what it is. I mean it does make sense, but it's like I worked so hard on making the app work and it turns out just like flipping the two prices is like maybe that's actually all you need to do.

[Josh Moore]

Yeah, it's funny, like some we middle always works. And then like, honestly, if you do rows, the first option is the one that most people pick anyways too.

[Steve P. Young]

Oh, interesting.

[Josh Moore]

It's really crazy. Yeah. And then one tip I'll give you too, is one of our clients, they tested Columns versus rows. And look, everything is different. But they noticed the columns where rows worked a little bit better than columns. So three rows versus where the annual. Probably the first one versus three columns.

[Steve P. Young]

Interesting.

[Josh Moore]

Might be worth.

[Steve P. Young]

You also told me of one of like trial could be like a toggle. I've started seeing that where you pick. Do you want a trial? Like, that's kind of interesting because I'd be like a no trial guy. I'd be like, no, I'm good. Let's just go. Even though it's like a free day, I don't know. Everyone's psychology is a little bit different.

[Josh Moore]

Yeah, I like it, man. That's awesome. Congrats. So Black Clay, I would say most of the time I like to do have a. Have a trial. We have seen that work better with. In a bunch of AI apps that we've worked with. The trial has worked. So most of the time I try to do have a trial. Especially if you have a hard paywall. You might want to consider this. Josh obviously gives away 30 minutes free every month. And Josh, I like every day, you said, or a month, you say every day. 30 minutes. Every day or 30 days?

[Steve P. Young]

30 minutes a month. Sorry.

[Josh Moore]

I love it. See, like, don't give away too much. Give them a little bit of taste and then they're going to have to pay. Like, you have, you have a business to run.

[Steve P. Young]

Yeah, well, yeah. And my, my intuition was that you try it once and you're like, oh my God, I got to buy it. And then I looked at data and so many people buy it before they try it. Like, they see the ad, they're like, I'm in, you know? Yeah. And so that's when I started showing the paywall right away. Like, even though you don't necessarily have to use it and it has a big X in the corner, but some people just see it and they buy it right away.

[Josh Moore]

I love it. I love it. Okay, cool. And then we got some questions for you. How did you manage to describe more than one hour audios?

[Steve P. Young]

Yeah, so I think. And then the follow up I could see in the chat was about the OpenAI API. I don't use the OpenAI API. I do use Whisper, but I host it myself. And so there's no limit. You know, this is a very good example of a place where I try to innovate right away. Like, I started on the OpenAI one, I started dividing the files and all this shenanigan. And I realized that actually running OpenAI on my own rented hardware would Be cheaper and better. I could do speakerization, which I'm about to launch, which is like Steve, da da da da, Josh, Steve, you know what I mean? Which the OpenAI API does not do natively. And also I don't have a limit because I realized people were going to be using this for long form stuff. And so you can record as many hours as you want in one shot. I personally wouldn't. If there are any people listening, if you have like a four hour thing you have to record, I would do it in one hour chunks just to get more data, keep it a little bit cleaner. But you can record as long as you want. And so that's how I got around that.

[Josh Moore]

All right, Ollie's here from London. Good to see you. And then Digihod said, would you mind sharing, Josh? Feel free.

[Steve P. Young]

All good, all good.

[Josh Moore]

All the.

[Steve P. Young]

So this one was not intuitive. Sam Altman, Uncle Sam has done me a solid because every time I start to want to reduce my cost of OpenAI usage, he cuts price by like half. So I was using GPT4 Turbo or maybe just GPT4 and it was some amount and then I was like looking at maybe using Llama, the open source language model. And then like right as I was ready to go, he dropped a big cut and then 4.0 had another cut and now there's these like mini models that are essentially free. So the OpenAI cost has only ended up being, you know, 2, 3, 5% of revenue. The reason I'm squishy on the number is because annual payments should be kind of accrued. See, I'm like new at apps, but I'm an expert in, you know, like accounting I guess. But you should take the annuals and kind of spread it out over the year and see what the cost losses for a year. So because of the high ratio of annuals that I have, I don't have a perfect answer for this yet. But it's, it's less than you think because it just, Sam Altman just keeps making it, you know, less and less. And I suspect that language models will converge on free over some time horizon. That's certainly where it's going now.

[Josh Moore]

You got a lot of questions coming in. All right.

[Steve P. Young]

Yeah, Rock and roll.

[Josh Moore]

This is this person Digihaud is. He loves you. Josh, what kind of events? And I already know this, but like. Well, I'm guessing based off what you already said, what kind of events and data do you send to Facebook Ads for it to find the relevant users?

[Steve P. Young]

Sure. So I put the Facebook SDK in my app, I have had MMP problem. I haven't been able to get an MMP in the mix to do better than just the Facebook SDK. It's. I'm only doing ads on meta. That's what I use. And they just send the subscription, you know, event, like this person bought. And it's all kind of automatic. It just sort of sits in the app and does its thing. So it's. It's been really. It's been really amazing. I mean, it just sort of like now I have someone who's running that part for me, so I, you know, and I believe he was a guest recently. So you should listen to that episode. But, yeah, that's how I do it.

[Josh Moore]

That's awesome. And then it's just to get a little technical. It's one event and then the values are sent. So, like, you know, whether they sign up for a yearly or a monthly or a weekly, it's one event that's being sent to that. All right. Like it's a subscription event or do you have to split up with.

[Steve P. Young]

No, it's just a subscription event. And I'm actually hitting the other window now just to peek at the code because I don't remember because my brain doesn't work. I mean, we're basically. We're sending a subscribe event, I don't think. And then. And then. So I use Adapti and I send some stuff through them. Yeah, I've seen. I send some stuff there. I'm not sure. I don't think we're sending the dollar value of the purchase, but I'm not sure.

[Josh Moore]

Okay.

[Steve P. Young]

We really should.

[Josh Moore]

Well, I just want to know from a technical standpoint that it's just one event, whether whatever they buy, it doesn't subscription.

[Steve P. Young]

Yep. Yes. Yes, that's right. Cool.

[Josh Moore]

I love it. Good work there. All right, let's get into some. I got my friend question says, great, great app, Josh. How did you balance getting free users and having them pay and having to pay for AI for them? Did you have a lot of iterations to find the right tier breakdown that 30 minute mark?

[Steve P. Young]

No, I mean, the 30 minute was a sort of a guess. I think I might have at one point had 60 minutes. I'd be perfectly honest, I don't remember how I got to 30. I do monitor how much usage every day is like new member usage versus a subscriber. Subscribers are the vast majority of it. If it got out of hand, I might do something about that, but it really isn't. I mean, I Think the value of what you record is rather immediate, and then it decreases in value quite rapidly. Like, what I mean is, it's a meeting you just had. You need this, you know, you want the summary of that. Then in six months, it might not be valuable anymore. So the usage kind of is repetitive, if that makes sense. And the OpenAI bills were. They were much more as a percentage of revenue in the beginning, but they've gone down quite a bit. Whereas now the advertising is far and away the largest piece.

[Josh Moore]

That's awesome. Congrats on that. And then I'm assuming maybe you can. I can expand on this with Mikel. It would be like, which tool SDK did you use to measure meta ads roi? It's just the meta SDK. And that's all you have in there. You don't have the MMP or anything like that?

[Steve P. Young]

Yeah, I've tried to make MMPs work, and I'm bought into the idea if people are coming from different sources and all this. But for whatever reason, the Facebook, you know, judging simply by running a campaign on meta that optimizes to an event from the Facebook SDK versus an mmp, the Facebook SDK does much better. And I think because I'm not rolling in fancy to the MMP offices, I'm not getting much attention from them. So I'm just like, okay, you know, I would like to give you money. If you do not want my money, that is okay.

[Josh Moore]

No, I love it. I love that you just have that. We're trying to do that with our app too. And I'm sorry to come around on this, Josh. That's why, like, this is very timely too, because part of me has always been, like, cheap, you know, like, let me hack everything. Let's not forget, let's like, let's not spend money. Let's just see. And then now I'm like, okay, we're gonna have to spend. Let's spend to make. So how long? I guess when you started Facebook, what did you have? I love the fact, and I want to ask about why you're just one man. But, like, did Market just take care of all the ad creatives? Were you creating them yourself? Like, talk to me about that, like, how you got started with the Facebook stuff.

[Steve P. Young]

Yeah, I have worked with a couple of different agencies who make the ads. I've learned in the last few months about things like the hook in the first few seconds to try to get someone to stop scrolling and just all these newish ideas. The way the ads are shaped my intuition on Ads is way off now because just of the way the world is now. So I, I have some folks who are good at it, doing it. You know, it's like the ad is what you'd imagine. It's like my boss was really impressed because now I'm taking these awesome, you know, things like that and, and that's, you know, the. I sort of feel like, yeah, you know, the, the demo has tended to be a little bit older. Like we're doing a lot in the 30 to 50 range, which makes sense. Those are people willing to pay for apps and this is useful for them. And maybe they're not, as I said, like bathing in all the AI stuff online. So they don't know and you know, so that this is like a useful tool for them. Not that someone who is well versed in AI would not find it useful. I think they would do, but that tends to be who. Who finds it. I forgot what the question was.

[Josh Moore]

No, and then just the. So you hired the agencies. It was like, how'd you get started in the ads? Were they running the ads or were they just creating the ads for you and then you were running them or were they just taking care of the.

[Steve P. Young]

Yeah, they did production. I trafficked in the beginning. Marcus Burke, who I referred to earlier, he's very busy with me. You cannot hire him. Everybody, but you should try anyway.

[Josh Moore]

I have another email saying, can you intro me? I'm like, dude, just reach out to him.

[Steve P. Young]

Yeah, that's what I did. I literally heard him on a podcast, I heard him on Sub Club and I was like, I love this guy. This guy's, this guy's speaking my language. And so I pinged him and I was like, yo, we should do some stuff. I think you'd have a good time. Let's go. And so we did. And yeah, so he does the trafficking. To be honest, I've only done it a couple times, but our settings are rather basic. Like we believe in the model, like the AI meta ad model where they are able to find your people. You're kind of renting that ability. So it's not the most sound long term strategy. I'd rather have a website that search ranks very well. I'd rather have people, you know, referring and talking about it. And I'm building all of that stuff too. But I mean the meta ad engine is the best version of that that humanity has ever known. And it's just honestly, like kind of amazing to me every day when I look at it.

[Josh Moore]

Yeah. Like, would you mind if I Show this your. Some of your ads. I'm on your thing. I'm on your face.

[Steve P. Young]

Yeah, let's look.

[Josh Moore]

Okay. I know everybody can do it too, but I was like, all right, let's.

[Steve P. Young]

Of course. Nothing. Everything someone wants to know, they can figure it out. I put enough info out there, they can get there.

[Josh Moore]

Thanks, John. You know what I really love? Oh, here, I'll play this real quick. I don't know if you guys can hear it.

[Steve P. Young]

We can't hear it, but it's. It's great. He's like. He's like, download this app. All I have to do is set it up before every single meeting and wave. AI will then transcribe and summarize all the conversations for me. So he thinks I'm doing all the grunt work. All I'm doing is sending the summaries that Wave generated for me. Check it out. Wave on the app store and 3x productivity. So that's definitely one angle. I think in that video, he's in a zoom call on his computer with the sound on and he leaves the phone on his desk. And that is not. That's a cool way to use it. That wasn't totally the intention, but it is a good illustration of kind of what it does. I think it's more exciting to do it, to have kind of real world recordings, but people do that as well. But these ads tend to work. And then those are some old ones at the bottom.

[Josh Moore]

I love it. Congrats, man. This is cool. Thanks for being so open. That's what I was going to say, Josh. Like, that's why I learned from you. You're like, how'd you get to 10k? You're like, I just kept sharing my success. Like, as I hit a certain benchmark, you kept sharing it. And then when I saw yourself on, like when I saw the posts on LinkedIn, I was like, I gotta reach out to them. And I was like, wait, I did talk to them. Yeah. It all comes after a while.

[Steve P. Young]

Yeah. Build. There are people who do this, like 100 times better than me. I. What I mean is my sharing of the graphs is like, build in public. Here's my graphs. Like, people are much more subtle and nuanced about it and much more entertaining. And they're sharing lessons and I would love to do that. I found that I'm not great at it. It's not that interesting. And I'm kind of busy. I'm like one guy doing this. And so I have not gone that route, but I sort of tried a Little bit in the beginning. And it reduced eventually to, like, showing pretty graphs. And that tends to get people's attention.

[Josh Moore]

Yeah.

[Steve P. Young]

Just to, like, get the sparks going a little bit. And frankly, I have no. I Very few people to, like, high five the successes with because the team is really just me and a couple of Me. A couple of guys.

[Josh Moore]

Yeah. Come on, let's go.

[Steve P. Young]

Boom.

[Josh Moore]

All right. We did the nerdiest thing possible. My wife's gonna.

[Steve P. Young]

Oh, I think we're gonna top that. I think we'll top that. Yeah.

[Josh Moore]

Oh, yeah.

[Steve P. Young]

I've seen these shows before. I know where they go.

[Josh Moore]

All right. Namit says. Hey, Josh, do you still run ASA Apple Search ads after starting Meta?

[Steve P. Young]

Yeah, we have a little. I think we've turned them down a little bit, but I. So. A little bit, but not very much. People, I think, are bidding on Wave as a name. That's always fun to see. I always click on those ads. Run up your bill for you there, buddy. But, yeah, a little less now.

[Josh Moore]

That's hilarious. And then this is a great question from Jananis. What was the biggest breakthrough in terms of acquisition? Was it just a single CRO test, a winning ad concept campaign structure app experiment that increased retention by a few multiples?

[Steve P. Young]

Yo, so I. I was watching an episode of this show, and you had a woman who was on. Yeah, man. I mean, if they're listening, they're already, like, into it. But there was a woman on who I thought was so good, and I think she was, like, making ads for games and she. And she drove home that, like, if you get a creative that works like, it's an. It can be an outlier difference. I've not yet had that experience. What I focus more on and what has moved the needle mostly negatively is, like, paywall issues. We had a build that the paywall didn't fire at the right time or it didn't fire at all. Or instead of 80% of people seeing it, only 40% saw it, you know, and those things depress revenue materially. So it's been more. What's really moved the needle is just making sure the app is tight, that people are seeing what they're supposed to see. We put an onboarding in relatively recently, like a few slides in the app the first time you open it. To sort of give you a sense of what it is. We've experimented with an aha moment type thing where you would like a guided test of Wave. Like, you pick the audio sample and you put it in and it does it. And I pulled that out I think it was confusing some people. One, like, nuance that I think, I don't know if it's normal, but it causes some trouble, is that I force making a user account. And the reason for that is I want your email address. But the real reason for that is I want to make sure all your recordings and data is secure and it's saved server side. And then if you go to wave on the web, which we have now, or an Android or on another device like your iPad, all your recording data is there. And I didn't want to tie those accounts to the Apple system. So the first time you open, you know, you're creating an account either with Google or Apple. You didn't say any audio, so it couldn't, it couldn't record anything. Oh, I can't hear you at all. I think you're muted, bro. My man Steve, you're muted.

[Josh Moore]

I'm not very professional.

[Steve P. Young]

Are you kidding that the like mute symbol was so fancy? I didn't even. I knew because it was like part of the graphic.

[Josh Moore]

I, I wanted to see when you asked for it. I don't think I ever signed up. I don't know if I got the.

[Steve P. Young]

You must have. I'm. If you, I don't know if you're on a phone. Yeah, I think you might have. But if you log. Because remember, we've spoken once in the past, so you might have already. This is the. Yeah. These are the screens you see the first time.

[Josh Moore]

Are you, are you a designer too? Like, how'd you come up?

[Steve P. Young]

I have a design firm that's, that's doing this for me.

[Josh Moore]

Got it.

[Steve P. Young]

Late checkout, name call. They're, they're really, they're really terrific.

[Josh Moore]

Late checkout. Is that Greg's late checkout?

[Steve P. Young]

Yep.

[Josh Moore]

Okay. Yeah.

[Steve P. Young]

Okay. So that's great. I'm really glad that the paywall showed instantly there. I mean, you had already opened the app on another account. But that, you know, the difference between that showing right away and not showing five seconds later is a huge difference. It's a huge difference. And so, you know, getting that right, getting the introductory processes running correctly. And remember, I've never built an app before. This is my first one. There's all sorts of nonsense that I do that doesn't make any sense that any like real engineer would be like, this is disgusting. But I've, but I've up leveled over the last year and so now it's like kind of reasonable.

[Josh Moore]

All right. People are asking some intimate questions. You already get intimate. Do I have an Intimate effect. I don't. Maybe this will work. Good enough. Yeah. Okay. All right. We're time to get into it. Yeah. All right. Vince says, what is your CAC?

[Steve P. Young]

I would love it to be like 30 bucks or maybe less and sometimes a little more and that's okay, but you know, as little as possible. But yeah, I mean I think, I think I buy at 30 all day.

[Josh Moore]

Yeah, yeah, of course. I mean we can easy more money than that. And then how much do you Tahor says, how much do you spend on ads daily?

[Steve P. Young]

We did 100k in July. You know, we're a very, we're a very workday heavy app. This is a funny story. So I mean we're obviously dead on Saturday, Sunday, like literally minus 90% of recording volume. Whereas on a given day like Yesterday we did 200,000 minutes. And so on a Saturday we'll do like 20,000 or 10,000, like really very, very little. But on the day that all the Windows machines all over the world broke like two Fridays ago, wave was also very low. Like people didn't go to work that day. I can kind of see in our app data when people are using the app or not, it almost feels like.

[Josh Moore]

A B2B play where it's like, you know, you're adding functionality. Obviously people are more willing to pay. You can charge a higher price.

[Steve P. Young]

It's for sure. It's a work thing for sure for most people.

[Josh Moore]

Yeah, I like it. All right. And then Vivek says and love your insights on this. Does charging more in developed countries than other countries developing underdeveloped increase overall revenues? Is it primarily US for nano?

[Steve P. Young]

It is primarily us. I definitely have like on my list of things I want to do is lowering the rates in certain places where it doesn't. It wouldn't feel appropriate if you saw it. I think Philippines, we were talking about doing a cut there. I think in Turkey prices already are low from a change. I all of a sudden, very early on, so many Turkey downloads, like hundreds. Like early on when I had no downloads, like Turkey became a number one country. I don't know what happened. Someone like wrote about it or something. But now it's mostly us and other English speaking places though it really does work quite well in any language. It'll grab your devices language and assume that any audio recording is done in that language and then do the transcriptions and summaries in that language as well. Early on there was a lot of German and Spanish and French and Portuguese and all this stuff that is now simmered a bit. It's just mostly a lot of English.

[Josh Moore]

All right, I love it. Okay, Leandro, now, last of these questions. Love from Turkey. What was your Google Play Web revenue? Yeah, go for it.

[Steve P. Young]

So we, we came out on Google Play last week and So I have 14 subscribers there and in the App Store I have about 10,000. And on the web I have implemented Stripe, but I haven't told anyone about it. So the only revenue is my test transactions. There's actually a bunch of features that are like built and ready to go, but they're not quite ready to go that I just start telling users about on the support channel, for example, people are like, it would be great if I can access my data on the web. And I'm like, well, you can go to App Wave Co and log in and see all your things. But it's like not pretty and it's not done, but you could go and do that. But it basically works. There's a bunch of things like that that I'm pushing out solely the Android app took forever to get live, not to code, it's the same, it's react native, so the code is the same. But I got rejected from the Google app store 40 times in a row before I was accepted for like little stupid things. Not even good things, more like, that button doesn't work, you idiot. And I'm like, oh yeah, my bad. And then that, that happened like literally 40 times. So that sucked. But now I'm live and it kind of works.

[Josh Moore]

Got a few cost questions and now you've been able to do it.

[Steve P. Young]

Let's do it.

[Josh Moore]

Could you expand more on hosting your own AI? Was it difficult? Any learnings we could benefit from?

[Steve P. Young]

Yeah, I mean, so whisper, the OpenAI transcription model, is open source, which means you can download it. You can download the model. It's been packaged up in some open source things like Whisper X and Faster Whisper, which are like mods essentially of Whisper. Then you can launch that. There's a lot of API services. RunPod who I use, are fantastic. They're more technical, but I wanted to get the cost as low as possible. If I recall correctly, the OpenAI hosted Whisper API is something like 36 cents an hour. And I think they, they, they call it six cents a minute. I'm sorry, they call it six tenths of a cent a minute or something like this, but it's 36 cents an hour. I can do it for like a quarter of that myself. And you're like a little bit on your own if things go sideways But a lot of this, a lot of what I'm doing here is like a project to learn how to do all this stuff. And so messing with Dockers and like I had heard about those things, I didn't really know what they were. But ChatGPT can really do every. It can, it can teach you anything, particularly technical things.

[Josh Moore]

Yeah, sounds like technical questions. Without an engineering background.

[Steve P. Young]

I was like, what is Docker? How does Docker work? And I'm like, okay, cool. Here's a bunch of docs from this Run Pod thing. How does this work? If you're clever about what you're asking, you can learn anything, particularly technical things. It's really just incredible. Things that, by the way, you could have learned with just Google, but it would have been harder. You'd have to find the right website and learn the information has been there. It's just packaged in such a better way as a chatbot.

[Josh Moore]

I don't know if you covered this because I'm not technical, but. Josh, could you share how do you manage token usage in your app? I guess it's one of the biggest pain points for fans.

[Steve P. Young]

So this is. So I was. So that question and others led me to sort of like intuit what a server is. I was like, I need a computer that can run some of these more private things like the OpenAI call. So I use Google functions, I use Firebase functions for all the back end stuff. Really all the app does is it records the audio and then uploads the audio. Once the audio is off your device, everything else happens on device. I'm sorry, everything else happens in the cloud. So transcription, summarization, saving it, anything you do, if you make it a favorite or put in a folder, all that stuff is recorded on Firebase on my database. This is an organic learning because what would happen, it wasn't about what you just asked whoever asked that about. The key is like the secrets you know, it was more about. You'd record the audio and send it up for a thing and it would take time for it to come back. And if you like changed apps or if you turned off your phone, it would just lose the recording. So I sort of figured out early on that I need to get that audio off the phone and do all the processing. That isn't going to be instant. It's going to take a few minutes of like waiting for the server to reply. I learned about webhooks and sort of asynchronous things like that. It has been my goal and I've been like, Adjacent to this I have a degree in math. Like I'm not a random. Like I'm an analytical guy. Anyway, so this was a bridge for me. I don't mean to imply that this is for everybody, but for me I was close to it already and so AI was able to get me the rest of the way there. But I learned about servers and how they work and cloud functions and serverless and all this stuff. It's like related it to. If you've ever been to. If you've ever been into a show like you've watched like a really good show like Breaking Bad or Mad Men, any of these great shows and you have to wait because the like season ends and you're waiting for the next season to start. Like how is it going to resolve if you just come in right now to become an engineer? There's all these things have already been built. Like it's amazing, like making websites on next js I was like, how can I like from the website like have a. Have something done that like makes an action happen on the server And I like look around like they just invented something called Server Actions. It's like terrific. Like I'm coming in so late to a lot of this stuff that it's all fully baked. I don't know if that made any sense. This isn't live, right? It's really.

[Josh Moore]

I love it. No, I love this. I love this, Josh. Honestly, it's just like to see you talk about this because the website's beautiful as well, man. Like it really is. And you're like, hey, yeah, yeah.

[Steve P. Young]

The website is actually the one piece of customer facing code that I didn't make. I asked. I hired a guy on upwork to implement the design because it has this cool scrolly thing that I just couldn't figure out. Someone asked a question about token usage.

[Josh Moore]

Yep.

[Steve P. Young]

I don't think that's next, but I just want to jump to that. So how do I minimize token usage from Russia? One of my instincts on this is that the thing I want to do is set no limits. I want to use the best APIs for the best summary, the best transcription. I don't want to hold back at all. Use all the tokens you want. I think it actually just makes the product better. Here's some free advice for anyone out there. I see this thing all the time in AI startups that the free version is using GPT 3.5 and the pro version is using GPT 4O. So you're using the better AI models for the People who pay. That is in some ways maybe even the opposite of what you should be doing. But it's certainly half wrong. Like, use the best models, give the best version of the thing in the demo. Like, it doesn't make any sense in my view, to make the demo less good. You should make the demo more good. There's so many things I do to like optimize for the new user for when they first download and they push the record button and they record. Hello, what does this thing do? Stop. Right. So they've recorded three seconds. When the average recording is 15 minutes, they've recorded three seconds. I want to make sure a transcription and summary come up immediately. Right. So I run like excess servers to just make sure there's no waiting, no cold starts, no queues, like you're going right away delivering that top notch experience. So with token usage, I have been, I'm trying to be as generous as I can for people who don't know what that means. When you run a command in AI, you're kind of the length of the response and the length of the input is measured and you're built based on that. That's how, that's what, that's what Russ is asking there. How do I minimize that?

[Josh Moore]

Love it. And Leandro, I'll answer this for Josh. Yes, he does provide multi language conversation support.

[Steve P. Young]

Yeah. Yep.

[Josh Moore]

All right, let's see. And I'm going to answer all these questions. Josh has a hard stop. And so I'll get to the app audits by myself without Josh. Unless you want to stay around. It's up to you, Josh.

[Steve P. Young]

I'll jump on. I have a few extra. I got a few extra minutes. I got a few extra.

[Josh Moore]

Okay. But I, I want to make sure, you know, the, the audience gets the, the questions answered. So what was your budget for the first marketing campaign?

[Steve P. Young]

Yeah, so I think above a minimum threshold, below which you're not going to learn anything like. So as long as it's above a certain amount, the budget is kind of infinity as long as it's able to meet its ROI goals. And so like practically speaking, I think I started with $300 a day. You know, I had, I had had the feeling that that would probably be enough to get enough installs to start to see something. But it's really like I'm managing spend. I'll spend double or I'll spend half what I spend right now based on what the results look like. That's really the key for me.

[Josh Moore]

Okay. I love it.

[Steve P. Young]

I hope that answers are you looking.

[Josh Moore]

At it on a daily basis too, like because you see the download, the activation, because you have no trial, makes math a little bit easier. Most people buy during the onboarding and so that's right, kind of almost.

[Steve P. Young]

Yep. So in Adaptee, who I use to manage the sales, they and this is going to answer Siddharth, they give me ltv. They have a really good data and analytics package on Adapti. And so one of the things that there's one graph where by day or by week it'll tell you when people subscribe, if they subscribe. So by zero day, 60% or 250 subscriptions, then three day, seven day, you know, 30 day, whatever. Because people will all install the app on day zero and then they might convert on a bunch of different days. So we look at that, you know, if you spend $100 on Monday, you want to. It's not just how many sales you make on Monday nor is it how many sales you make that week, it's how many sales Monday's group makes in that week. Right, that makes sense. Like you want to follow the cohort. So we look at the cohorts, we've been able to generalize certain things. Like if we see X installs or subscriptions, really subscriptions on day zero, we can assume a trickling of subscriptions from that group over the next few days and you can model getting ROI. Like basically we want payback of more than 1 on that first sale on like unlike a, you know, a blended basis.

[Josh Moore]

Have you tried other channels or are you just using Facebook?

[Steve P. Young]

Tick Tock was sort of my very first experiments. I stopped doing that. I found their dashboard a little bit difficult to navigate. You know, I think if Meta felt like it was tapping out, I would start exploring aggressively some other things in terms of like scale. But Meta touches really everybody who uses a phone between all their apps. And so I think there's enough there for now. I definitely want to try to do other. Well for Android I need to do Google Ads at some point. I haven't started that yet. And they're like little one offs, like inclusions in newsletters that might be a hit, kind of less scalable, lower dollar. But just like good opportunities. Like I sort of think opportunistically about that. But I personally what I, what I love doing of all this stuff the most is the engineering work. So I'm trying to keep everything else off my plate and really just focus on that.

[Josh Moore]

Yep, got it. I love it. Would you still do the same thing? Like if you had to do another 0 to 1. Would you wait till the 10,000 mr? Would you know what you already know now? Would you start with Facebook ads right away?

[Steve P. Young]

I think like it sort of depends. I. So when I was thinking about it, about this, like I have the money to lose, but that's a dangerous place to be in. So I wanted the project to be self sufficient. Like that was kind of the idea net of like a credit card that will give me a 30 day rotating balance, you know what I mean? Like I put the money on a card, I get the money from Apple, they're due and I get the money at around the same time in a month. Right. So I was willing to use that float, but I didn't want to go terribly in the red. I think I seeded the business with some amount. Like, I don't know some amount. But then I really wanted it to be at worst, break even just as a, like, you know, just as like a rule because it could easily get carried away. So I think that's a, that's a reasonable rule. Like it gets dangerous when you try to make your acquisition equal LTV, you know, like well LTV equals $100. So I can acquire users for a hundred dollars and it'll pay back in the first year. Like that's one way to do it, I guess. But as an indie it feels dangerous and it wasn't what I didn't want to be like that.

[Josh Moore]

Well, like what you said, like I have money to lose. But that's a dangerous proposition too because yeah, right.

[Steve P. Young]

Like, yeah, the market will speak. I mean we didn't talk much about this, but even at Uber, like we were pretty, particularly in the early, like before we were worth a billion dollars. Like money was kept very tight. Like, you know what I mean? Hiring was a big deal, spending money was a big deal. Like you got to keep it tight. I think one of the problems in venture backed startups, that's not really what we're talking about here, but there's a lot of money sloshing around and companies get undisciplined because of that. You know, it's becomes not a big deal to spend 100 grand, but it's a huge deal to spend 100 grand. If you have 20 million in the bank or nothing in the bank, it's still 100 grand. It's a lot of money. So you really need to make sure that the economics make sense because you turn up the heat. Because the fact is like if your app works then the economics on the ads should also work a little bit. And if it doesn't, there's a signal there.

[Josh Moore]

Yep. Love it. That's love it said. All right. Siddhartha said has to. He asked it twice. Okay. What is the best ratio for new user to paid acquisition? Paid conversion.

[Steve P. Young]

So we convert at 10%. I suspect that that's on the high end, but I have no idea.

[Josh Moore]

It's pretty high. It's great. I say 5 to 10 is on average, and then above 10 is great.

[Steve P. Young]

Thanks.

[Josh Moore]

So. Yeah, right. Josh, people are asking. It's dad jokes. Nure. It feels funny if you call it daddy jokes. But we'll get into the dad jokes in the app audit, and then we'll let Josh go as well. All right, let's get to it.

[Steve P. Young]

Okay. Me first.

[Josh Moore]

Yes, let's.

[Steve P. Young]

All right, first, why did the orchestra get struck by lightning?

[Josh Moore]

Why?

[Steve P. Young]

Because it had. Because it had a conductor.

[Josh Moore]

Oh, shoot. It was too low. I like it. All right, I gotta see if I can try to eat that. All right. Hey, Josh. My wife gets mad over my bread puns. My wife get mad. Gets mad at me over my bread puns. Why? She sourdough.

[Steve P. Young]

Oh.

[Josh Moore]

All right, so put. Jay, if you.

[Steve P. Young]

Hold on. One more. I got one more for you. Well, I got one more for you. What do you call. What do you call a fake dad?

[Josh Moore]

What?

[Steve P. Young]

A faux pas.

[Josh Moore]

All right, that's two. You guys are killing me.

[Steve P. Young]

You're killing me here.

[Josh Moore]

If you thought Josh's joke was better, put Jay and then put S. If you thought my joke was better too, we'll play for bragging rights or coffee drink. We'll figure.

[Steve P. Young]

Oh, I didn't know this is gonna be. I didn't know we were doing that. All right. Hey, thanks, guys. Give me those J's. Give me those J's. Pass that J. Let's go. All right, let's get in. What else you got?

[Josh Moore]

This app. So you got time for this or you wanna. Do you need to go? Because I want to make sure we ask.

[Steve P. Young]

Do it.

[Josh Moore]

Okay, so Leonette Leonardo is here. He wants store meta metadata and then onboarding. Why don't we focus more on the onboarding side of things? I think the metadata looks pretty good. Kids Storybook. Truth. I don't know if this is a good word to put in here in andro, because it's. I don't think anybody is searching for your brand, so. I like the brand in here. I like Kids Storybook. But you might want to go into, like, kids bedtime stories or something of that realm. That Might be more ASO friendly here, but that's what I would say for here. And then he does have like children's book. Let me pull up his app. Let me stop sharing a little bit. And then come on, pull up some different screens on my. So here is the app. He's got children, books, bedtime stories. Yeah. Right here. So I might even just lean. Whatever has the most traffic lean into that term in the title. And then you do have. So the ASO wise, having the main keywords to the left is probably. It's better from that. All right, Josh, you want to. Anything that you want to give feedback on for the ASO side of things.

[Steve P. Young]

The rating is good. I've noticed that Android ratings are so. Are so much lower for whatever reason. So I respect the 47 on the Android Store. I'm an Android noob, so I don't. I don't think I have anything for you.

[Josh Moore]

All right, let's get into the app. I think that's where we can make more because he's got 50,000 downloads. So like that's great. Got your app in the background. Feels slow. Just like what you remind me. He's like, hey, five second difference. So watching a video, which is great. What is this? Mace mes? No, feels like an error. Start my free week. Feels expensive to me, especially on Android. You might want to play around with the pricing here. And Andro, what do you think, Josh? Like would you felt. What'd you feel about this whole onboarding?

[Steve P. Young]

I love the animations. I mean, it's really nice. I'm not. Like I said, I'm not. I'm kind of new to Android. What. What's the like. Is this. Is this U S based? What is the 54.90 per NES? Yeah.

[Josh Moore]

I don't know. I'm assuming that's a year. Yeah. But I don't know what the MES.

[Steve P. Young]

Is, so you probably want to have a language. I love that. My kids apps always have that. Turns out they just learned what year I was born in and now they get past it.

[Josh Moore]

It's hilarious. Yeah. It looks like for the first year, it's 50. It feels expensive to me.

[Steve P. Young]

Interesting.

[Josh Moore]

Yeah. I've got some experience with some kids apps, so, yeah, it feels a little expensive. You might want to play around with the pricing, especially on Android. It's. Oh, he's here. Android's here. It's $55 a year. So play around with the pricing. Leon. Andro. Because we're doing that with our app that we just launched. And we think it's probably a little bit too expensive. We did get some sales already, but this is on iOS typically Android doesn't convert as well as iOS and so people will tend to even have lower prices on the Android side versus the iOS side, like 10, $15 cheaper on a. On Android. So that's what I would say. Anything here, Josh, you want to add?

[Steve P. Young]

No, I think you got it.

[Josh Moore]

Let's move on. I do like the paywall, the start free trial, all that stuff. Yeah, everything else looks pretty good. If you find valuable, please set for review. Oh, okay. I mean you might want to put leave a review or write a review versus leave a comment. I didn't know what your comment.

[Steve P. Young]

Yeah, I agree because you're, you're sort of asking for a favor there. So you should, you should be clear.

[Josh Moore]

Otherwise I like how you had Intercom in the early versions. I don't know if you still do. Is that something you would recommend to a lot of the app for something?

[Steve P. Young]

Yeah, I mean it's such an, it's a, it's a no brainer. I think everyone, you know, it's, it's cheap for the first couple years. They have a startup thing. I feel like the price is going to get really high when I finish that. But it's kind of a superpower to be able to let users talk to you right away. And I actually have a thing on iOS that I stole from someone else. But when you hard press the app icon such to you want to like delete it, I have a shortcut in there. Like chat with me, you know, like, do you need help? Like talk to me right now. And let me tell you, that literally blows up my phone wherever I am. So it's not totally scalable anymore, but I still do it.

[Josh Moore]

I love it.

[Steve P. Young]

You know, just being a human. I've always found and this was true ever. I've worked in e commerce and apps and all this. But when you screw something up, if you can fix it, like really fix it, you can almost be better off than you were if you didn't screw up at all. So.

[Josh Moore]

Okay. Love it.

[Steve P. Young]

Yeah, I think it's, it's, it used to be a little easier to get to me now messages. Yeah. But just like good to be available. You know, people are always, people are always surprised. Like, wow, you're like, yeah man, I'm the engineer of the app too. Like you're not just talking to someone. Like I'm actually going to fix your problem. That goes a long way.

[Josh Moore]

I love it. Well said, my friend.

[Steve P. Young]

Thanks.

[Josh Moore]

Okay. And that's why I would recommend, too, because I think, you know, from my lesson with Josh, Josh here doing this interview, you might even Leandro, like, you might even want to try a discount or like, not giving any type of trials. Yes. Like, most people will buy, but I have a sneaking suspicion because I was in the kids app, like, kids apps. As a parent myself, I just want to make sure my kids like it. I'm willing to pay. I just want to make sure they like it. So I'm almost like, hey, I'm okay with them trying out a few stories and then if they're very engaged. All right. You know, like, so maybe you don't need a trial. You might want to play around with the pricing and you might not need a trial for this one, too. So I might want to play around with that, especially on the Android side, because Josh has no trial and he's doing pretty well.

[Steve P. Young]

There's a trial, just not using the trial mechanism. You know what I mean?

[Josh Moore]

Right.

[Steve P. Young]

It's a. It's a freemium, as it were.

[Josh Moore]

Yep. I like it. Cool.

[Steve P. Young]

Thanks.

[Josh Moore]

Russ says, can I ask one more question? How many organic downloads do you get? I mean, percentage of total downloads.

[Steve P. Young]

Yeah. Well, this is one problem. I don't. I don't have a great answer to that, unfortunately. And this is where the, like, MMP piece comes in. Because the Facebook stuff is so large scale, you know, I know it's more than zero, but it's certainly a concern that I don't have a great answer to that right now. Do you have any best practices for knowing? Yeah, if I were to stop ads. Whatever is left over.

[Josh Moore]

Exactly. And like, when you run Facebook, obviously your brand search volume goes up. People are just going to search for wave AI in the app stores too. And so we see a natural trend. The line there. There's a strong correlation between paid ads and app store search downloads because there's always going to be a section of the audience. So I'm glad you said it in the early days. In the. Earlier in the interview, you said, hey, I look at the blended, and I think that's the way to go, because we do see, like, all downloads. So if you're just running Facebook, look at your all downloads, because then it'll give you an indication of how well Facebook is actually performing.

[Steve P. Young]

100%. And I'm. And I'm. And I definitely buy in to that. That someone sees an ad, maybe they search for it later. Sometimes, though, I have, like, I'll do An appearance, you know, like this. I've done some things like this before that I know, drive a bunch of downloads and you see a bump, but it's hard to tie. Like event based, organic, you know what I mean? Like you did something. And so it got a bunch of inbound. Having a kind of a hard time measuring that. But it's not a priority at the moment.

[Josh Moore]

For the people that create and manage your ads, do you give them a cut of the profits or. Depends on the ad spend? Percentage of the ad spend?

[Steve P. Young]

That is a good question. It is a percentage of the ad spend.

[Josh Moore]

Okay.

[Steve P. Young]

And I think, you know, there. Yeah, you know, it's still kind of a new thing. I think there is. You know, that's a traditional ad agency way to do it. In many cases, they would also make the ads if they were doing that. I think it's. It's. That's how I do it now. I don't know how it'll be long term.

[Josh Moore]

Yep. Love it. Okay, how about. How are you in time?

[Steve P. Young]

I gotta go in a couple minutes because I have another one coming up. Not one of these.

[Josh Moore]

Why don't we do this then we'll let you go and then I'll take everything. You did win. So I owe you something, John, but thank you so much. You got three to one. The app is called Wave and you go to Wave Co and then if you search for Wave AI in the app store, you'll find it there as well. But Wave Co and then Wave AI in the app stores. And Josh, if the audience wants to connect with you in any other way after this, do you want to send them anywhere else?

[Steve P. Young]

Hit me on Twitter, Oshmore. I answer basically everything. So send me a question. I will most likely answer it. Thank you all very much. I appreciate it.

[Josh Moore]

We got over 100 people live tuning in. So thank you so much for sharing your insights, man. Congrats on all success.

[Steve P. Young]

Thanks very much, Steve. Have a good one.

[Josh Moore]

Keep in touch, man. Cool. All right, I'll get into Alex's app a little bit and we'll go from there. So great question. And I'll try to get to. I've got a hard stop in a few minutes too, but I will get to it. That was a great interview with Josh. All right, Room cleaner, Alex's app right here. Let me pull up his app store. He just wants onboarding and first and after onboarding, user experience. So, Alex, I'll take a look at it for those who are just listening. It's a storage cleaner app. It Removes duplicate photos, videos, so forth. We've worked with a few different apps like this too. So I'll get into the details of this, Alex. All right, let's get into it. I'll allow you to track. Welcome to Broom. Love it. I'm trying to swipe, by the way, Alex, and I can't swipe, so maybe think about adding that. Like, I'm just trying to swipe my phone like this, but you're forcing me to hit. Get started. Yep. Full access. Okay. Okay. Love this. It's just analyzing my storage. Love this. And then three days trial. Wow. It's super cheap. Okay, cool. Yeah, I love this paywall. I mean, there's nothing. Okay, you have all the little tricks here too. I like this. I think I would play around with the price. You know, what we found is just for your lifetime. It's for those who are just listening. It's 6.99amonth, which I like. You might want to play around a week. Maybe people just want a week. So maybe you can get away with just a week versus a month. But that's just feedback. I don't care. If you change that, then 20 bucks a year, three day trial. I love that too. You might want to play with not having a trial here too, especially after Josh's interview here. And then lifetime. I'd make it 29.99 because we found that these type of weird numbers, like 30, 99, don't convert as well as just showing a usual rounded number, 29.99. So put that test out with that because I think you'll do very well with this. Especially with. I Love this. Lifetime. 7.99, but crossed out. It worked really, really well. So I love this. Alex. This is really good. Nice. Let me just delete some photos because I can. Okay. Delete. Selected. I like it. Oh, love it. Okay. Love this. Delete. Selected.

[Steve P. Young]

Delete.

[Josh Moore]

Delete. Selected. Let's play around some AI apps. Okay, let me just say delete all. This is where you can start asking for reviews, Alex, because I just had a good experience with your app. So think about asking for a review. And then even after, if I leave a five star review, I mean, I know you can't tell, but I would try to show the paywall more often. Hey, you know, want to delete more? And again, like, I would limit. Just sort of like play around with what Josh said to the 30 minute. Like give people a taste of what they can delete. Don't let them do get away with everything free. Because most things we will get away with. If we can get away with things, doing things for free, we'll do it. So here, let's go. Oh, I like the swipe mode. Delete to keep. This definitely. Delete, keep, keep. Of course. This is my Tinder profile, right. I'm just kidding. Cool. Trash. Yep. Delete. Yeah. Again, here you can easily share. Get them to rate. This is what I would change. Yeah. Cool. You. I think you waited too long. There you go. I got. I saw the review and then show the. So paywall again because look, they've had a good experience, right? They're more like. They're probably more likely to pay. Want to free up more of your stuff? The other option is, as I do that, give them a discount. Hey, want to free up more space? You can save, you know, five bucks off the lifetime. So now the lifetime is like very close to the yearly. Like as I'm using the app, that's a way for you to convert because, you know, ultimately they might forget you come back in. Just leave your app again. So I would play around with the. The monetization here. Let's see if you show your paywall. Yeah, you do. I would probably around with not having a trial. Like, I know I promote having a trial, but I think I like your. The way you've priced everything that I would probably try not having a trial there too. Okay, cool. That is it. We had a great live stream here, guys. Let me see if there's any questions. Thanks, guys. Thanks, Josh or Irvin. Where you been? Will this interview be recorded? Campbell, I just saw this in my YouTube feed. Thanks for doing this interview, Steve. Thanks for joining. It's always going to be live on YouTube. All right. It's always live. People ask. This is always live on YouTube. And then it goes. If you just want to listen to the audio only, just go. Go on the podcast every Monday morning. Right? So. Oh, okay. Thank you guys for answering it. Okay, cool. I think this is. I'm using. I mean, Ben says I'm using Remy Cat right now, but paywall and it's pretty weak. But at least I have some testing. Is software to test opening journeys. Oh, Ben, you might want to check out Adapti for onboarding stuff too. Adaptee has for that. The revenue Cat. I love them. I think they're really great for analytics. They're not strong on the paywall side of things yet. And so look at Adapti. Look at Superwall for some of the paywall stuff to do things to like, do a lot of like ab testing. And like changing the paywalls on the fly without needing any build. That's what those things would be. All right, Alex says thanks for the feedback. Pricing strategy is a bit old. Yeah, it's a great app team, I think. You know, let me just see your reviews if you're not getting that many downloads yet, Alex, you know the lifetime offer for free on app advice, that should hack a lot of your ASO and your downloads there too. So. All right, this was a great stuff. Thank you. See in the community. All right, you there, Ben. I know if you're in the community. Okay, Abdul, I've got. I'll create a quote, another video for this, but I've already taken a screenshot of this. Which tools would you recommend for identifying relevant keywords before building an mvp? Yes, great question. I'm going to create a separate video off of this app follow. Right. This is what I would use. You just want keywords. So app follow to do keyword research. It's not working right now, but let's just assume like note taker. Right. For Josh, I put that in here. AI notetaker. So he's number two. This is great. Now, app follow will work eventually, but you hit this little eye icon and this is how I do keyword research and you get the popularity. It's unfortunately not working. And then secondly, Abdul, what you would do is put it all those keywords to app radar to get the difficulty scores and then launch the app that way too. So that's how you. That's how you do it. Those are the tools. Steve, please tell me what to do. Ratings. No showing. I have no idea. I know you asked it a long time ago. Steve, why Apple not showing, rated or review on app? I'm sure five, six people did great, but not showing it because I put. Yeah, look, Norex, I think that's what's happening right now with the iOS. Apple's like either not showing it as much. Look, that's why I'm like, honestly, like, sometimes I share too much information. I feel like this is. You guys. Tell me because you guys are in the community as well. When I share it, does Apple start figuring out to look for it and start minimizing it? This is not a common theme. This is a common theme. Now I'm shared showing the review prompt early on. Now a few years later, people are like, oh, maybe I'm not doing this anymore. I shared the whole discounting during onboarding and we had a bunch of different rejections. Now that people are doing that, like more Frequently. So I feel like we have to be careful what we share publicly because Apple is paying attention. All right. But that's just me in very little things. All right, Leon. Andro says, any tips, references for landscape apps, mostly games. We've had a hard time finding good references. You know what? Leandro. So especially in the space, because that's what you are in the. Look at, I think Lingokids, what they do really well is it is a landscape app, but the onboarding is portrait mode mode. So the onboarding is vertical, which is natural. And then as you get into the app itself, paywall, everything's vertical until you get to the homepage of the app and then that's where they show they go landscape mode. So look at Lingokids. You know, when you guys ask questions like this, that's when I really feel like I know a lot about apps. All right, cool. Everything else. Think we're good in right now. All right, that is it, guys. Thank you guys for showing up.

[Steve P. Young]

We.

[Josh Moore]

We hit over 100. Hey, if anybody has any contact at Twitter, please let me know because we go over 100 all the time. When I was streaming this on my personal Twitter profile, but unfortunately you need to be premium. And if for some reason I can't sign up for premium on my personal Twitter profile, so this goes live on the app Masters, which a little bit less following still is doing really well. So thank you guys for tuning in. Next week we're going to have 10. Jen, come on. And we're going to talk all about, like, advanced UA strategies and analysis. So look, I do think that. And we're making moves on our end to help you guys with UA and Facebook ads. And so I feel like I haven't talked about this publicly because I wanted to feel really, really confident that we're really, really good and help you perform. And so we're going to talk all about how do you analyze your UA efforts? Because as much ASO I like to talk about, as you can see with Josh, it takes you away too. ASO is important, but not the end all, be all anymore. So I want to really focus more on the UA side of things and really getting good about that. And I feel like we're there. And so we're gonna. We've been providing UA and Facebook ads for a lot of our clients. If you want to work with us, go check it out. Just reach out to [e-mail] and we've got a video editor, we've got a whole production team to handle your UA efforts. So go to appmasters.com if you want help. And we're going to be running it on my own apps, too. That's what next week is going to be all about. How do you analyze your UA effort and really figure it out? Because that is a very complicated situation as well. All right, that's it. Babbling on. Thank you guys for joining. Every Friday, 9:00am Pacific. I'm back, baby. Content's been really good, right? I think so. All right, have a good weekend, everybody. I will see you every week.

[Steve P. Young]

Thanks for listening to the App Masters podcast.

[Josh Moore]

For show notes and amazing app marketing.

[Steve P. Young]

Content, check out Appmasters.