3 $1M AI startup ideas targeting profitable niches and how to build viral apps

13.5k views6373 WordsCopy TextShare
Greg Isenberg
Join me as I chat with Blake Anderson, Co-Founder of Apex, as we discuss his frameworks and strategi...
Video Transcript:
so what's up Blake what's going on Greg thank you very much for having me on today I decided to reach out to you I've been following you for a while um but I I saw a tweet about a month ago that said I recently discovered a part of the internet where a bunch of cracked 18 to 22 year old Indie hacker kids from noname Unis run boot Trap mobile apps growth hacked on Tik Tok each app has a 100,000 monthly users and makes $15 million a year never underestimate the kids and did some research and they were talking about you and your friends and how you have kind of like a Ai mobile app Studio doing I think you know 15 million in ARR so I was hoping that you can come on today and share with everyone some AI app ideas yep happy to do so start [Music] quick context on myself I kind of stumbled into this uh space a little bit over a year ago with an app called Riz GPT it is now named plug AI uh we end up scaling that to a little bit over 3 million downloads 200,000 monthly Revenue then built umax uh that has done over 10 million downloads 500,000 monthly Revenue now I think it's closer to 400,000 but at at our Peak we were around 700,000 in monthly Revenue um and then most recently Cal AI which we launched just like a little bit over 6 months ago and is at about 2 million downloads and over a million dollars a monthly Revenue yeah I think I'm uh like I mentioned I kind of just happened to stumble into this space but it's a pretty fascinating place of the internet and specific specifically entrepreneurship to be right now because it kind of takes the intersection of new functionality in terms of applications that can be built as a result of AI crossed with novel distribution tactics largely pioneered by uh e-commerce people um and so when you take the two of those together it creates for this uh this space where you have like a small proportion of kids essentially like age 18 to 22 I myself I'm 23 uh that are kind of running the entire uh the entire space and at the Forefront of it absolutely man and I think uh a lot of people like me who are in my 30s and I'm sure people listening in their 40s 50s 60s are kind of like what's going on uh with the Tik Tok kids and um it's interesting to see that a lot of the products have ai with it I mean I'm not surprised but um yeah what's your take on that like is there an opportunity before you get into your ideas and before we like start cooking basically is there an opportunity to create apps that are have like not a or do you have to come up with an AI app idea for Tik Tok yeah so I believe that and I'm really excited to get into some of the ideas that we have to share uh later on but I believe that we're currently just scratching the surface of new uh application functionality that can be built and delivered to the end user um using these new AI tools so for example uh with umax UMX was a facial analysis and recommendation application in the past this was possible to build however it would have taken significant investment into developing internal prop proprietary uh ml facial recognition models in order to build it however with the use of AI uh specifically GPT Vision I was able to do so with essentially zero upfront development cost with the exception of my own time um I think that what we're seeing right now as I mentioned we're like scratching the surface of these new functionalities but what AI really enables us to do is to deliver personalized and like personalized functionality that can be updated in real time um through the use of these AI backends as well as building uh technologies that previously would have costed a lot of money and so we're seeing the new Tik Tock AI apps like this is the start but as we progress as AI becomes uh more advanced and uh increasing functionality as as well as people building more complex uh applications with it I think what we're going to see is just a 10x um in N user experience uh with regard to how we interface with technology cool well yeah let's let's dive in what's idea number one what do you got for us so as you guys have heard the the applications that I built have really been uh around sort of AI virality uh the first two specifically R GPT how to text girls Max how to look Max right but Cal AI was that was primarily around functionality you take an archaic uh sort of Industry uh calorie tracking where it was not very easy for the end user ton of people have gripes with apps like my fitness pal and you make it easier with the use of AI and so this is what I increasingly think about not building apps where you can just kind of uh you know leverage the Tik Tok algorithm to hack your way into a little bit of cash but like how how do we redefine how the the value that people get out of so my first idea that I have is uh an application for lack of a better word career AI as we all know there are a lot of new very popular um fragmented career advancement tools by fragmented what I mean is you have a website that you can go on that gives you an analysis of your resume and it'll give you all the B bullet points about how to improve it we also have websites where you can take quizzes in terms of uh that will give you recommendations of what your future career should be based on your personality we have people that sell courses about how to apply to different jobs but I believe that an all-in-one integrated platform that kind of takes a bunch of these different fragmented um uh tools and puts it into career AI I think that that would be a pretty amazing application uh and specifically as someone who just recently graduated from college I know the trouble that a lot of people have in uh finding jobs finding internships that sort of thing um and it's a massive industry to begin with like one of the uh primary selling points of universities is that like they're going to set you up for your future career now I think that a career AI you charge 50 bucks a month for an all-in-one Tool uh to me feels pretty obvious um and then with regards to distribution I would go about it you know the pretty standard Instagram TI to reals obviously Target students uh but then you could leverage Twitter or x. com in addition to LinkedIn for some relatively untapped uh distribution channels shh don't tell anyone but I've got 30 plus startup ideas that could make you millions and I'm giving them away for free these aren't just random guesses they're validated Concepts from entrepreneurs who built hundred million plus businesses I've compiled them into one simple database compiled from hundreds of conversations I've had on my podcast but the main thing is most of these ideas don't need a single investor some cost nothing to start I'm pretty much handing you a cheat sheet the idea bank is your startup shortcut just click below to get access your next cash flow business is waiting for you so I want to go deeper into basically how how you have like what's the Blake Anderson framework for for taking an idea and actually making it like a mainstream product I guess um so for example like you said yeah just Target students with Instagram reals and you know LinkedIn and Tik Tok like could you walk us through what you would do if you were building this and you had this idea like okay you have the idea where do you go from here yeah so I believe that there is a uh people often fall victim to seeing the new uh trendy way to go viral and then putting all of their eggs into that basket I think that something that all of my applications have really benefited from is I'm engaging in a highly iterative process of generating virality from early on and so that means coming up with new hooks figuring out new ways to position the product in terms of who you're marketing to um and split testing those over a rolling basis so let's say that I really think that influencer marketing is going to work with uh this career AI application I'll find influencers in the General Ed Tech space uh that talk about study tips I'll find influencers that give tips on how to apply to jobs I'll get find influencers that give General productivity tips I will come up with different hooks that could uh theoretically work within each of those niches I'll then split test these hooks across different videos across different creators and assess the results assess how many views we're generating based on which hook uh what the relative conversion is based on on a per viiew basis uh for each of them and then over time you start to develop either a mental or real statistic model statistical model um that will be able to inform your future uh distribution uh strategy I hope that makes sense but somewhat akin to how platform platforms meta Tik Tok Etc will split test your different creatives uh applying the same strategy to influencers or internal ugc or uh you know creative hook generation that sort of thing and how do you find the influencers like do you do you go through um do you just like kind of reach out to them yeah so I was actually looking at this today because I thought that this is uh pretty indicative of what I do which is kind of different from a lot of people number one I don't use uh influencer marketplaces or databases number two I think that there's a lot of messaging on the internet about various tools so like the marketplace and databases or agencies um to to find influence for you that's just like outright wrong what I don't see people talking about is the development of personal relationships with influencers so I just recently launched my new Company Apex I can get into that a little bit later but in short it's a very content heavy company we're creating content in-house and now you'll see I make this post and like I guess this is backwards but I've got a bunch of creators like Sam Zia commented he was a former co-founder from Max 220,000 followers Dawson Gibbs 170 or 170,000 followers back State Carter 280,000 followers the list goes on and on um and now the reason that they're commenting on this isn't because I'm paying them to but it's because I've developed the relationships with them and I think that or I know that these influencers are getting hundreds of DMS emails what have you on a daily basis it's really hard for them to differentiate what companies are serious and what companies aren't but if you can embed yourself uh and develop personal relationships with them embed yourself within the industry then they'll connect you with whoever you need to get connected with so yeah I mean it's it's it's funny you mention it like to me it's like you say this and it's obvious you're basically like yeah you know they're human beings uh you know create a relationship with them get to know them better and that's the that's kind of like the most effective way of doing it um um but yeah that's kind of like the a bit of the opposite of what other people say which is they're like yeah go on to these platforms reach out to you know what I mean listen I I totally agree with it reach out to as many as you can I always say prioritize early on both quantity and quality you got to reach out to thousands of them you gota you gotta if there's a creator that you really want you have to DM them on every single platform you have to message their mom you have to message their girlfriend saying I want to pay your son money please put in contact like I've done that before that like that can kind of help solve your cold start from there you got to treat them well become friends with them and then the opportunities become endless the product like we kind of you know we said it's an all-in-one product but like how do you how do you take an idea how do you take this idea and actually flesh it out and and make it a real product like how do you think about it so okay first off I think that similar to what I was saying about split testing different uh hooks across different niches about why people should download your product if for example with career AI here if we're embedding multiple different tools into the platform what we can do is we can also split test different product hooks so we could test let's say we have three different tools within the application then we could test different tools on social media to see what's going to go the most viral and we could engineer around virality with umax for example this was the uh facial analysis ratings we found that this sort of six Factor Madden rating system that's what people were most drawn to that's what people were most likely to like comment engage with and therefore went the most viral I want I want to to make that clear from the start because that is kind of how I begin my ideation process what are the tools that are going to do well on social media explain see so so that will inform design and then my process for design is usually I do wireframes I have either a professional designer or someone on upwork convert that into more flesh out UI uh and then it becomes time to build with regards to building I myself I use chat GPT to teach me uh how to code to build Riz GPT and umax now if you have Capital this is not a route that I recommend if you have Capital connections to Great CTO like don't you don't need to do that yourself but if you're just getting started like I was at the time you get the designs done and then you start you figure it out so for me it was talking to chat GPT on a daily basis uh originally I used Swift UI so I was getting code from chipt and Swift UI placing that into xcode moving forward um now when I code I'm doing it in a cursor with react native I just started doing this about a month ago quick note I did not code CI so I coded R GPT in the first version of umax yeah I think that people are often looking for like a one-size fits-all answer when it comes to the development side of things I just don't think that it exists I think that it's different for everybody butt what I say is what I'm doing now is a react native incursor and with career AI specifically like if you were building this like what features do you think would be need to have so I would think around uh two things like think around the cross between functionality and virality uh the resumé analysis that's that's very high functionality now how do we take something that's functional right scanning someone's resumee and giving tips on how they can improve it how do we also make that Viral so what I would probably do here is a sixf factor rating system similar to umax I would give them scores based on different sections of their resume now I would see design those scores around the sort of popular edte content right now getting cont with thech creators and then have them start to post so that's just on the resume analysis side of things I would also do a career analysis you take a quiz and then at the end of it it's a nice pretty screen that says you should go into X Y and Z and then I would probably figure out how to create a trend which is like people being really excited because they're a senior and they got recommended computer science and that's what they've been studying and then people being like wait why did it why why did it take until I'm a senior to realize that I shouldn't be studying this and then create create that content on social media and I think that I can make that Viral relatively quickly cool all right thanks for that free idea career AI what's uh idea number two okay idea number two is Journal AI I haven't seen anyone crack this quite yet I know people have spoken about it the way that I would build it is somewhat of an AI second brain but branded and framed as a journal um and the sort of analysis that you would get would be things that would help increase your productivity emotional support relationship analysis um and I really don't think it would have to be that technically robust I would say essentially just chat GPT uh with uh more complex memory storage and UI that's more conducive to journaling so when I hear that kind of like maybe because I've been around for a long a while like there's so many journaling apps is your your thesis is that the AI angle is enough to kind of to get people excited is that is that the thesis yeah so I think one of my thesis here is that I've seen people organically use chat gbt as their journal and get a lot of value out of it I've essentially done that myself um now I'm not a daily journaler but when I provide chat GPT with uh you know I have like I have notes on my computer that just have a lot of information about some topic and uh maybe too much to be contained within chat gbt's memory and when I want to go get a deeper analysis I'll go paste those in and now I think that uh yeah sort of like an emotional support assistant framed as a journal would be pretty easy to to get it to go viral as well as I think that there would be pretty high value ad um you know we have this like one-sided relationship with journals where we put in and get nothing out and that's great to some effects uh but I think the ability to get that sort of like intricate objective analysis back out similar to how we like I think a lot of people building this are thinking of it as like a therapist AI but I wouldn't brand it as that I would brand it as a journal AI if that makes sense yeah I remember Peter levels I'm sure you know him he he did AI therapist. com or something and he he he closed it down also like there's like you can't just call it a therapist because like therapy is like you know uh you can't just it's not a therapist like you know you have to go to school and stuff like that to be a to to be a therapist so um journaling also the other interesting thing is it's something that people know like they know what a journal is so if you call it like AI journal or Journal AI I mean that's a question for you like how do you think about naming some of these products and is putting and we're really getting into the nitty-gritty but I think it's important is putting AI in the name important yeah people ask this question a lot I think that uh it's hard to say at this point in time I think that the the uh appending AI is valuable when you are creating Innovation within a relatively stagnant industry such as Cal AI within uh the calorie tracking space like calorie tracking as a whole people viewed it as archaic and it's slow and it's annoying and it's complicated so then when we add AI that brings a level of sophistication now on the flip side when it's something where you know if if you're building like a medical uh doc doent service or something where like you need 100% accuracy um and any minor errors uh would would be unacceptable the system just wouldn't work I would probably not include the term AI as people would be a little bit skeptical or scar that it may become dangerous in some capacity uh so so that's how I think about but you know I mean my co-founder Zach from CI recently made a tweet about how adding AI at the end was a negative signal he and I had a lengthy discussion on that kind of came to this resolution that I just referred to just now um but yeah I guess I guess that's how I think through it and now with regards to journal AI um the the the names that I'm using right now are just because they're easy to wrap your head around I'm not sure if I would name it Journal AI maybe I would try to name it something like juny or like Journey that like feels more friendly um but yeah I wouldn't say that I'm by any means a master of naming so maybe they're better people to go to on that front yeah but I think all your names resonate like there's a reason like if you're my my thesis is that if Cal was just called calorie tracker it might not be like it probably wouldn't be where it is today that's my that's I agree I I think C is probably the name I'm most proud of um the reason for naming umax for example umax was because uh the plan was to become this like all-in-one self-improvement ecosystem that I'm now building with Apex now for a number of reasons I decided not to pursue that but in retrospective umax if the plan was to take it to where it is now I probably would have named it looksmax AI as our copycat ended up doing um so quick ad break let me tell you about a business I invested in it's called boring marketing.
com so a few years ago I met this group of people that were some of the best SEO experts in the world they were behind getting some of the biggest companies found on Google and the secret sauce is they've got a set of technology and AI that could help you outrank your competition so for my own businesses I wanted that I didn't want to have to rely on Mark Zuckerberg I didn't want to depend on ads to drive customer to my businesses I wanted to rank high in Google that's why I like SEO and that's why I use boring marketing. com and that's why I invested in it they're so confident in their approach that they offer a 30-day Sprint with 100% money back guarantee who does that nowadays so check it out highly recommend boring marketing. com okay cool and let's just talk before we move to the next idea can we talk about how you think about pay walls and are there any tools that you use uh for testing pay walls and and stuff like that and and how do you think about just like how pricing strategy Shameless PL for this for this in particular too like this idea yeah uh Shameless plug super wall absolutely amazing I am I'm not affiliated I just I love the platform I think that pretty much everybody should be using super wall I don't use Revenue cap I only use super wall now with regards to how I think about pay walls I think it really depends um so what's amazing about super wall is it makes split testing incredibly easy they provide you with all the data but I think that across the industry a lot of people tend to optimize for like number of users multiplied by profit per user I am of an alternative thesis I believe that you should optimize for like number of us users raised to the N where n is greater than one multiplied by profit per user um I will take at least at a just pure conversion basis I will accept a lower LTV on a per download basis um in exchange for more people being able to use the platform now there are a number of reasons for this but I think one of the um most obvious if you want to ignore all like the moral like oh I want more people to use the product more people get value of it if you want to completely set that aside um which I don't but if you were to set that aside the benefit here is that you can more quickly dominate the market and as a result um increase Word of Mouth which is pretty much impossible to measure uh for most applications uh it's easy to measure like if you're just seating a thousand users onto your product and then letting it grow organically but when you're constantly running ads you don't know if people are coming specifically like organic you have no idea where people are coming from um and so I like to prioritize allowing as many people to use it as possible now um moving forward how do I think about pay Walling in itself like within the application super wall is amazing once again because it enables you to test it at different points and you can assess the relative effects the conversion basis how many people are seeing it that sort of thing um the most standard practice in the industry right now is to take people through some onboard earning flow and then hit them with a pay wall at the end um it's just like you make so much more doing that than pretty much any other pay wall strategy when you when you say that are you talking about like it's a quiz or something like that and then at the end it's like pay push notifications and put your personal information so we can customize the application for you and then you get hit with the pay wall um I've done this for my applications for the most part uh I do not necessarily think that it's great practice but it is efficient practice if that makes sense like I don't think that it's morally great I don't like feel good about it but totally cuz like people are investing like they they might see a Tik Tok about it or real whatever they download the app they're getting they're super excited they're filling in their information they're investing essentially their time yeah and some and then they're just like oh $6.
99 a week or something yeah yeah I think yeah I don't know I'd like to believe I guess it depends on how you do it right yeah I mean I so yeah I don't I don't think I I don't I don't think that it moves the needle much like I don't think that it's that good or bad I'd like to believe that with what I'm doing with the money at the very bottom of the funnel like what I'm using it for with what I'm currently building uh kind of offsets not a big deal uh like I said I think that this industry as a whole doesn't really move the needle very much uh but it is you you know the way that I generally think about these things is like capitalism is good in my belief and therefore increasing market efficiency in uh generally just taking the market in a quicker manner to its inevitable end State um and so yeah I guess that's kind of how I like offload the responsibility dude we we have to do one more idea we have to give us one last idea so I think that this is one of the biggest applications on the App Store du lingo Duolingo is an amazing application like it's built out so well now that said if you can acquire even just a couple percent of the market share by integrating new AI functionality quicker than them you build a massive product and that's where I would uh say that a good idea is a language AI or language tutoring application uh obviously the chat GPT voice too model is currently pretty expensive but my inclination is that that will significantly decrease in cost over time uh or other models will come out of similar functionality with decrease cost so by building a language tutoring application where you are uh just always up to date with the newest AI functionality and you build the UI around uh around sort of of dynamic learning like as we know du lingo generally has more static modules that you go through ABC you build a language learning application that updates dynamically I think that there's a lot of value there um a lot of value to be created as well as a lot of value to be captured people who want to learn languages are generally pretty high intent so I think that that's a that'd be a pretty awesome one to build what do you say to people who say but dualingo just announced two AI features I think it's called like Adventures or something like that like how am I gonna how how you g to compete with dualingo yeah this is exactly what people said to me when we were Building C AI people are like dude my fitness pal so big they're going to build in this space and my answer was well they won't work as hard as us their incentives internally are not aligned to the point where they will be able to iterate and and progress as quickly as us so here's what I mean by that the engineers at my fitness pal like my fitness pal to begin with has been like traded back and forth sold so many times um that the people that are doing the groundw work don't like probably don't have that much Equity you know and so the engineers don't care that much like they're just kind of doing their job similar with the social media managers and so our CIS uh uh so calorie estimations based on images are just significantly better than my fitness palette we have a pretty robust system built out at this point in addition we're pioneering new ways of marketing on social media I would wager that the same pretty much applies to Du lingo now I wouldn't be surprised if their engineering team was a lot better uh significantly higher valuation now that said uh du lingo is the go-to application for language learning right now but if you can figure out just like a few percent changes to to for people to learn languages in different ways and then some novel distribution strategies for that do a lingo is worth like what1 billion if you can uh like I said take just 1% of that market share you immediately have an application worth $100 million and I think that for the same reason that du lingo is so big there's so so few people try in which just makes it so much easier to succeed uh when applying uh these sort of New Wave tactics yeah I mean the short answer is you don't need to reinvent the wheel you know I was actually earlier this morning I was on Tik Tok Creator insights and for people who don't know uh go on Tik Tok search for T Tik Tock Creator insights and you basically get like a Google Trends but for Tik Tok and you can see you know how much what are people searching for and what's trending and stuff like that and one of the interesting things was today what was trending a ton of how to learn French how to learn Spanish how to learn Arabic how to learn Arabic I remember search volumes up a th% so yes du lingo uh you know may continue to grow may continue to be a juggernaut and that's cool but you know there's an opportunity right now with um creating AI first apps that I think um when you pair that with great Creator content there's just an opportunity to not even take part of their market share because the the reality is these people are not even using dualingo that's the that's that's the reality yeah creating new market share I totally agree I think that that is a to dive a little a little bit deeper into something I mentioned earlier like I think that a lot of these apps are just pioneering new ways uh for people to learn and improve themselves with the use of AI I think it's a pretty imp incredible personalized learning tool and when it first Chad GPD first came out in like December November of 2022 there was a lot of speculation about the future of AI tutors and within AI or learning languages or personal training that sort of thing my inclination is that uh at the very minimum the first step of this is just AI tools AI apps that do this as opposed to being like a framed as a tutor it's just do it is doing the tutoring through its UI if that makes sense yeah of course of course so yeah I I think it's it's really awesome like the plenty of room for du lingo to coexist with an application that uses AI to Pioneer completely novel um functionality within the language learning space same thing goes with calorie tracking same thing goes with uh career advancement same thing goes with journaling before before we head out like do you have any like I think people listening this are like okay those are three ideas um but they might be like how you know I want to come up with more AI app ideas and then I want to create a framework I want to framework for that do you use things like sensor Tower do you use things like um like any tools like that or are you just coming are you just coming up with these ideas while you're walking in the beach in San Diego yeah so I certainly use sensor Tower to assess like the size of the um other players in the space but I think few things have helped me one as you mentioned walking the beach like for for me it's generally hiking I find that I'm able to think very clearly two deliberate scrolling on social media and curating my feed around whatever it is that I'm interested in so I'll create burner accounts to like when I was building umax I had a burner account where it was like only looks maxing content or anytime I build a new application I create a burner account and I curate my entire feed around that and I like put myself in the mind of someone who's consuming that content um and then finally I'm actually creating like a 30 to 60 Minute um completely free somewhat of a course or YouTube video that I'm going to be posting in about a month uh as well as a completely free uh viral app Playbook PDF and so if people want to you know follow my socials Blake Anderson W on Twitter or x.
Copyright © 2024. Made with ♥ in London by YTScribe.com