He makes $100k/month with his Lovable Agency

0:00 hey Harry great to have you on the No Code  MBA YouTube channel um Harry is the founder   0:06 of Imaginary Space an agency uh building  with lovable making real revenue and today   0:13 we're going to talk about the agency how much  money Harry's making building lovable apps   0:19 uh the the challenges that he's coming across  and a lot more so really excited to have you   0:23 here Harry thanks Surf hi everyone it's good to  be here harry Roper i'm an a AI consultant for   0:30 the enterprise industry we've been building with  Lovable for the last year and just seen like crazy   0:35 growth using this tool so I'm excited to to share  as much as possible well I guess just to to start   0:42 for for people who who don't know you or maybe  just to get people excited how much revenue are   0:49 you currently making building with Lovable yeah  so uh as it stands we're doing about 100,000 a   0:55 month um and we're on track to do 1.2 million  uh by the end of the year but that could grow   1:02 and it's probably the most success we've ever  seen with our agency honestly it's uh just been   1:08 crazy the last few months the amount of growth and  been able to grow the team from that and it's it's   1:13 just really exciting uh space to be in honestly  that's amazing and and what kind of apps are you   1:21 building with Lovable for clients um is there a  specific type or does it just run on variety of   1:27 things yeah we have like two veins of like clients  one is enterprise and then one is startups so   1:32 like some of the apps we're building are like  more about BTOC so it's like we're building an   1:38 edtech platform in Dubai right now for a um a  really big Arabic speaking school uh and then   1:44 on the other hand we're working with Dude Wipes  which is kind of like a household name in the US   1:49 um building like internal tools for them to help  speed up their processes so kind of spread across   1:54 the two we're still figuring out what we want to  do like longterm but mainly it's this idea of like   1:59 how can you build a rapid prototype of something  in like 30 days uh that's kind of our like uh   2:04 offer we have on the site right now and just like  getting product into people's hands so that they   2:10 can start giving feedback and and building upon  that fast uh we like to think that in comparison   2:15 to like normal agencies we're we're a lot faster  and a lot cheaper thanks to all of these AI tools   2:24 got it uh cool yeah I'm on your site right now so  the idea and is that what most people are reaching   2:29 out to you about they like that message and they  want to say "Okay let's let's build something   2:33 quickly and more affordably." And that's kind of  the the message that people are resonating with   2:38 yeah because I think the basis of it is is like  we've all built or like I I say we've all built   2:42 but a lot of people have built software projects  in the past and they've had like the same teething   2:47 problems like either it takes 6 months to get a  product out or it's going to cost them $120,000   2:53 to build it uh whereas like we all know now with  these new tools Bolt Replay Cursor Vibe Coding   2:59 that you can like 10x your output and thanks  to that uh we're able to offer clients like   3:05 this quick turnaround and it it seems to me that  like uh as long as you can make sure you still   3:10 hit quality in those kind of turnaround times like  it you can really make a big impact for a business   3:15 uh especially if like if it's a startup founder  and they're like okay I want to test an idea and   3:20 get out there quickly and figure out if it works  or you Right right to totally because it's about   3:27 at the end of the day it's about what what they're  getting doing it more quickly is is better also   3:32 i mean everybody loves that right you tell them  you can do something quicker they're like "Okay   3:36 let's sign up." Right right and are are you only  using Lovable or using other AI tools how are you   3:46 how what's your workflow there yeah so like just  talking about stack it's lovable supabase and then   3:53 sevalla for like deployment um and then cursor for  like small little code edits but I like to think   3:59 that we probably do like 80 to 90% of our work  in lovable and probably like 80 to 90% of the   4:06 code base is written by AI as well um crazy time  that we're in yeah yeah it it is and so sometimes   4:15 so you'll take the codebase and maybe do some  edits in cursor maybe some human edits a small   4:20 amount but it's it's mostly that local editor  yeah and then the agent's really good in cursor   4:26 so nine times out of 10 you're prompting that  to make the edits as well so like in regards to   4:31 like your audience and the no code MBA like it's  probably the perfect fit because you don't have   4:35 to write any code at the end of the day you just  do your thing right totally yeah and and that's   4:40 that's a similar workflow to what I've been using  lately is starting in bolt or lovable um a lot of   4:46 lovable lately and then yeah so maybe moving it  onto uh cursor wind surf just for some and using   4:55 the agent like you said so it's it's pretty much  the same workflow as lovable just on slightly more   5:00 complex platform uh yeah yeah yeah and do people  so so people are reaching out to you um do they do   5:10 they know that it's built in lovable like do they  is that what they want like have they seen you and   5:15 they they see that you go in lovable and that's  why they're reaching out to you um or what's that   5:20 what's that conversation like with with clients  i think like one of the benefits that we've done   5:26 one of the initiatives that we've kind of pushed  the last year is just making as much content and   5:30 relationships as we can and that seems to work  really well in terms of people seeing that we   5:35 know what we're talking about and then being like  okay I trust this guy you know people watch your   5:41 YouTube videos I'm sure and they trust you um and  any other content creators so that's really helped   5:46 and then what's interesting actually Seth is  that some of these founders are coming to us and   5:51 they've already built like their own mockup front  end in lovable and they've kind of put together   5:56 their ideas and they can show us like what they  kind of want but they just they struggle with how   6:01 do I connect everything how does everything work  together what are like some of the trends in UI   6:06 that I should be following what are some like best  practices so like we find a lot of the time now   6:11 we're we're meeting these really cool founders who  have done a lot of work already and they just need   6:16 help with like that last 60% you know but they've  done a lot of prep work beforehand um and then on   6:22 the other hand some people have never heard of  lovable but you know it's uh I used to be like a   6:27 web flow developer before and like what you would  do when you were selling the new technology is you   6:32 you show it so you know just showing like how easy  it is to run a prompt or like create a project   6:37 um and nine times out of 10 though it it's mainly  us that are saying like we should use Lovable and   6:43 that's just like obviously personal preference but  also we know how good the technology is so like   6:48 we'd be doing our clients a bit of a disservice if  we said don't use it you know right right totally   6:54 and at the end of the day the final product is  code it's it's not like you're even selling it's   6:59 it seems like it's maybe even easier than selling  something like bubble or a different no code tool   7:04 where um you're locked in because you're not just  have code and you can do anything that you want   7:09 with it afterwards yeah and I mean you've touched  it i'm sure you've had like loads of aha moments   7:14 using like lovable and bolt and and just seen  like the benefits right to totally yeah 100%   7:21 um yeah and and Okay so so I think that one of  the things that that people are thinking about   7:29 now that with lovable and and these and tools  like this are you know they've tried it out uh   7:35 they've probably seen some people building apps  with it the vibe coding trend of of who know this   7:40 is amazing you can build stuff but I think that  where people get overwhelmed or or skeptical is   7:49 can you actually build how can you actually turn  this into production apps that handle security   7:54 that handle all the things that that a company  would need to to actually you know put something   8:00 in into the hands of users so I' I'd love to hear  more about your process with Axe obviously as an   8:05 agency that is is it has to be more more than just  like a simple thing it has to actually be a full   8:12 full working app so um yeah curious if you could  speak to that and kind of talk about process there   8:19 yeah so for us uh because we work with a lot of  like financial companies we're always um making   8:26 sure that we understand the underlying technology  i think a lot of people and the mistakes that   8:31 they can make are that they don't understand how  Supabase works they don't understand how RLS works   8:36 which is Supabase's version of a file uh to put it  simply and I think like just for people to even if   8:44 they don't want to understand the code right just  understand the platform and like how things piece   8:48 together will add so much to your process uh I  talk a lot about it but like some people you know   8:56 will be using lovable or other tools and maybe  they'll deploy some stuff or make some changes and   9:01 they're not really sure what the agent is doing  i think just having a general overview of like   9:06 what it's doing and how it's working and putting  stuff together really helps um and then you know   9:11 just having a list of like the things you want in  your app that you want to build you know like we   9:16 are humans and we are creative and sometimes we  can go into this process of like doing lots it's   9:22 like a huge mess afterwards and um you haven't you  you've gone one direction then you go another you   9:28 know and like we always try and make sure of our  clients that before we even start building we just   9:33 have like a super clear picture of like where are  we trying to go and what are we trying to build   9:36 and what are like the five core things that like  you know it needs uh at the end of the day right   9:43 yeah and and at the end of the day like there's  no inherent reason why lovable can't build a you   9:50 know a production app it really it's writing  code it's just a matter of the person using it   9:55 understanding the technology like you said and if  you don't understand it then yeah you might leave   10:00 an API key exposed or something like that but  there's no reason that's going to happen with any   10:04 inexperienced developer um whether it's AI or not  AI yeah and like I think a lot of people question   10:11 as to like whether Lovable can hold up we launched  a startup last year called renameify.co uh which   10:17 got top 10 on Product Hunt and we had 7,000 users  on the launch day and nothing broke thankfully   10:24 um and like I feel like maybe some people are  hesitant but Superbase is a super strong tool uh   10:31 Lovable does write like pretty decent code nothing  amazing so I wouldn't be scared if anybody's   10:37 watching this like you know just make sure you  know what the AI is doing in a simple sense   10:42 and is your background in software development or  are you come from a nontechnical background where   10:47 are you coming from yeah I guess I was like a  developer when I was like 16 to 19 years old just   10:53 like playing games and like hacking stuff got a  job as a developer um writing a bit of code but it   10:59 was never really my thing so I went into website  design and I was more of a designer uh then I   11:04 got into automations and started playing around  like these no code platforms to do automations   11:08 for clients and then really like over the last  year of using Lovable and the exposure to stuff   11:14 I found that I've learned React and I've learned  SQL and like it's just been really helpful to be   11:20 able to read code even if I don't know how to like  write it you know from scratch at the very least I   11:26 can kind of read what's going on and be like okay  this is kind of the right way or like oh you've   11:31 made that mistake there and just like helping the  assistant you know as as need be yeah makes sense   11:38 i think that's a good er like that background of  just like building software whether it's like no   11:45 code or code or kind of having that going in i  think that ties in to kind of exactly we were   11:50 talking about before just like understanding  how apps are built like how are they how how is   11:56 a database connects to the front end the back end  all these things and I I've also found similarly   12:02 that like the more I'm building with lovable like  I'm like just learning how how it works like how   12:09 do how do the folders like set up how does the  code kind of work together even though I can't   12:14 write it myself yeah and that's it right just like  understanding how it's piecing it together but I   12:19 think Lovable's long-term goal is at some point  we won't even need to be developers in some sense   12:24 like hopefully the AI will just be able to take  care of everything and you know massage us through   12:29 the process and that'll be it you know yeah yeah  so so that that leads me into another thought of   12:36 where so so where do you find the AI shroules  now or or and are there areas of like certain   12:45 API connections or certain uh whatever it is where  like you start to run into issues and like how do   12:52 you overcome that with AI coding that you found  so yeah I think it's that basis of like building   12:58 really big apps a problem with AI at the moment  is they only have so much context that they can   13:04 take in and spit out so you find as you get to  these really big projects that sometimes it might   13:10 struggle to edit stuff there's also like that uh  piece on the AI just deleting stuff randomly and I   13:16 think it's taking these small steps and being very  precise with your changes that will help you a lot   13:23 of people will start a lovable project and like  write a huge prompt and like a huge requirements   13:29 file and I think it's much better to stick to this  idea of like small steps small concise edits and   13:36 just try and keep your changes to like a very thin  vin because unfortunately right now AI isn't like   13:42 the superpower we want it to be uh and it can't  just spin up like an app in one prompt you need   13:47 to kind of take it slow um and that's what I try  and teach my students a lot like can you just like   13:53 think a little bit more like sequentially and and  step by step in that case got it and and and what   14:00 do you do if you're in a situation where like it  just seems stuck like that you're you're trying   14:06 over and over to be as simple as possible but it's  not working like what do you do at that point as a   14:10 developer why combinator released a really cool  podcast the other day about uh AI tools and how   14:16 they're using VI coding at the moment and I  really like this idea of r-rolling so like   14:21 uh it's that idea of like if you're stuck just  restoring to a previous checkpoint where it was   14:26 working and then trying again and that's really  uncommon in the development world because usually   14:31 you're sort of coding coding code and you're going  to fix the bug but now because it takes a couple   14:35 of seconds to like spin up another version of  what you're doing it's really easy just to like   14:40 hit that restore checkpoint go back to when it was  working reroll again maybe change your approach   14:45 and and just don't feel like you're stuck uh or go  down the rabbit hole you know like so many people   14:52 cry about wasting credits with lower and stuff but  I bet if they just like followed that process of   14:57 being able to let go and give up and try again you  know they'll succeed a lot more yeah yeah that's   15:03 great advice um yeah don't keep trying to like do  something that's not working try start over which   15:09 is probably good advice for life in general yeah  hell yeah and I think you drive yourself a bit   15:13 crazy if you're like on a bug for like five hours  when you know you could have just rerolled it and   15:17 maybe fixed it you know right yeah that makes  makes complete sense something else I'm curious   15:22 so I I'm assuming you have a bit of a team now you  have quite a few clients you're working with um at   15:28 the same time Lovable is is quite a new tool so  is this something where like you're working with   15:34 people and then training them internally on how to  build are you finding developers and training them   15:40 are you um finding people who already are doing  this on their own like how what does that look   15:45 like from an agency perspective on your side yeah  great question i think uh we still try to hire   15:51 like traditional front-end developers just because  as I said it's good if you know how to recode and   15:56 it's good if you can jump into the project and  fix a bug with cursor or something like that a   16:00 bit more manually uh but we're really focusing  on this idea and I think Sav Sahul Lang from   16:05 Gumroad kind of thought about it it's like design  engineers so it's like guys who know or girls who   16:11 know how to code and they know how to like how to  design a little bit um so that they can produce   16:17 something beautiful um and what we tend to find  is like when you give a developer this tool first   16:23 there's the aha moment of like "Oh my god this  is sick it's speeding up everything but also   16:27 like after a week or two they kind of know how  to use the platform and we can share our our tips   16:33 and like prompts that work and stuff and and try  and help them out but I find that most developers   16:38 they get it pretty quickly honestly and they seem  to do well." Yeah got it that makes sense so just   16:44 hire developers who already know how it all works  and then okay now let's use this tool that's going   16:48 to make you work 10 times more quickly exotic  yeah yeah cool that that makes sense um and at   16:57 this point are you doing any developing or are  you kind of taking it more of a business side   17:02 of things in the agency um yeah I'm very grateful  that the team can like help me out i'm trying to   17:07 focus on growing and doing sales and all of that  stuff but I still do a further development maybe   17:13 on like the more complex projects um and I find  it's really good to have that knowledge like if   17:19 you ever walk into a sales meeting and you can be  like technical and also sell like it really helps   17:24 you know you can talk directly to their questions  you don't need somebody else uh to answer them for   17:29 you and you can kind of take the take the lead  on that a little bit yeah yeah that that makes   17:35 complete sense um well this is getting me getting  me excited to to keep building in in Lable oh yeah   17:42 oh dude so yeah and I feel like something just a  thought that came to me like if you're watching   17:47 this video and you want to get started with like  even just on a small scale doing some freelancing   17:53 with lovable like I feel like the first step is  just build a cool project for yourself and then   17:58 you have that just like what you said like what's  the best way to prove get a client like show them   18:04 what you built show them how cool it is and how  how powerful it can be and and I feel like it's   18:10 it's it's more accessible than ever to people to  to get started with this stuff and that's kind   18:16 of like how I started really i was using the  project in alpha last uh March or February and   18:22 was just building lots of projects for myself  and then happened to show it to like a couple   18:26 of clients that I already had or like new people  who were getting in touch and they saw the benefit   18:31 and the value in it and they were instantly ready  to work and it's like I think a lot of people try   18:35 and sell before they prove that they can build  something uh I can say this a lot to everybody   18:40 who's watching but if you build it they will come  um and you know if you post on Twitter every day   18:46 a lovable app that you've made or a little  project at some point somebody will pick that   18:50 up and say "Hey I need some help you know." Yeah  that's another great strategy just like go on go   18:55 on Twitter every day or every week just say "Hey  new project new project new project." Yeah you're   19:01 going to get clients from that hell yeah hell yeah  it's a lot better than just saying like cold DMing   19:07 i've seen these strategies where you just like  cold DM 100,000 people but it's like where's the   19:12 value that you can provide there you know right  so so where do you see the future of this how   19:18 big can this agency get like what do you what are  your goals um and and moving forward yeah I think   19:26 we would love to just carry on with the mantle  that we're the number one lovable agency in the   19:31 world um and continuing to work with clients i  think the interesting part for us is figuring   19:37 out like where is the sweet spot for lovable in  terms of clients and which people get the most   19:41 like which clients get the most value out of it is  it enterprise is it um is it these other kind of   19:48 companies but like you know I've seen other soft  traditional software development agencies do tens   19:54 of millions in revenue uh rather than just just a  million so for me the sky's is the limit and as it   20:00 continues to grow as it continues to get a buzz  I think the people who learn and become experts   20:05 in these AI tools are going to see the benefit  and the developers and people who don't you know   20:11 jump on this wave right now I think they're going  to fall behind um so you know for people who are   20:16 watching if you can learn these tools I think  it's a great advantage for you uh in the future   20:21 you know yeah yeah it makes makes complete sense  and what is one thing running an agency that most   20:31 people don't realize like if you've never run an  agency before what's something that that people   20:35 might not understand always turn up to the meeting  with a presentation and have a plan because a lot   20:43 of the time people don't and you need to make sure  that your client feels like they're looked after   20:48 they know where you're at in schedule and where  they should give feedback and I think my success   20:53 has always been that every single meeting there's  a presentation to run a client through and maybe a   20:58 lot of people don't realize that but when you just  turn up to a call and you're just asking questions   21:02 you're like "So guys what do you think?" You know  it's a little bit awkward but if you have a really   21:06 tight presentation that you can take them from a  start to an end point and then the meeting can end   21:11 your clients feel like they're in good hands and  and they're being looked after and I think that's   21:14 probably been the difference between me closing  a couple of thousand dollars to like a couple   21:18 hundred thousand in their deal cool yeah super  actionable yeah anything else you want to share   21:25 about yourself or the agency or where people can  find you if they want to want to learn more yeah   21:31 so Harry Mai on all social platforms i've also  got a course on lovable if you want to try that   21:36 out it's a classwithharry.com um if you want to  learn a little bit about some of the things I've   21:41 learned over the last year and uh yeah on YouTube  as well Harry Mckay Roper we're doing I do loads   21:46 of live streams with the Lovable team so go follow  them uh check them out and um yeah follow Seth   21:51 as well uh and the no and the no code MBA cool  awesome great well it's great to have you on Harry   21:57 and uh yeah excited to see see how everything  grows over the next year or so thank you thanks