Website available here: MakeLogoAI.com
This article is one of the next parts of my journey of learning to code and being profitable within a year
I was going to sleep but then “Hey, what if I made a micro-app to generate logos in 48 hours max”?
That’s exactly what I’m going to do. Bye sleep.
Here is the plan:
- Technical tests to see if I can do it (train custom AI model + output HD pictures usable for logos)
- Make manual tests
- Make a landing page with Typeform
- Make a post on Twitter
It’s probably not going to work as it’s a basic thing but whatever, that’s a nice challenge.
First step: Make a dataset of good-looking logos. I’m just going to check whatever I can find online.
Didn’t took long. Not perfect but good enough for a test. Launched 2 trainings.
First results are mitigated.
It’s definitely a start, but it’s not super consistant.
Did additional tests, the results are apeshit. Damn, this is gonna be more complex that I thought.
Still not as good as I wanted. I think I got the “HD” part covered (turn the logo into something usable for the web) but the initial design need improvement.
One last quick test and I’m going to bed.
Lol “quick test”.
Well, I found another way to make nice logos. It’s not what I planned at the begining but it will do the trick in the meantime I imagine. Bed, and tomorrow, let’s test the workflow.
Alright, answered all Twitter comments, took shower, protein shake, ready to work. Just realized the MVP is a direct competitor of dagorenouf’s Logology, so I hope he will take it well, I appreciate him 😅
Additional tests, the quality seems better! Let’s now try to generate everything programmatically.
1h I’m trying to convert a curl I found into an Axios function, nice.
IT WORKED! After some tweaking using chatGTP I was able to get it.
I’m now able to generate a full logo in .svg 🤓
Another whole fking hour spent trying to convert TWO LINES OF PYTHON INTO NODE.JS WHY IS THIS SO HARD??? GTP chat can’t even do it, ffs
Fuck that I’m moving to something else. So annoying, gosh.
So, I have a basic system in place to convert startup ideas (“An app that does XXX”) to logos.
Now, I either need to make a proper interface (can take more than a day?) or do everything manually on Typeform. I have 24 hours left so I’m going to make an interface to handle the results myself and then I’ll send an email.
Nice, was able to make a basic interface where people can enter their startup concept and it will automatically turn it into a logo and display it.
K, took some time to add better previews, now going to eat
Eat done. For the workflow, I think the best would be:
- Write your startup name
- Generate 5 concepts
- Pick the one you want in HD as upsell
Lol, seems like a small earthquake just happened in the coworking. No time to go out outside tho.
Took some time to think of the business model, so let’s implement it.
Let’s try that!
The “pipeline” seems ready, I can generate 5 icons quickly. Let’s make a quick display interface, then a “select” interface for the upsells.
Here is the flow:
- Landing (CTA)
- Write what your startup does
- Choose a package (Standard V.S Pro)
- Logo list
Gonna make a basic user interface. Will try to release the full version for tomorrow but worst case I have everything working on the back end, so I can make a manual version in the meantime. Gonna eat a peanut butter snack to take a break, will allow me to think.
Back to it!
Added authentification, going to use the MagicAuth like OpenForAds.com, to make a seamless user experience. Just have to copy/paste stuffs. That’s why I love making lots of projects, then you have pre-made templates ready to go.
Getting a bit overloaded as I have to go fast, so I’m going to slow down and write the steps clearly.
Alright, I wrote down the stuff I needed.
I’m going to use Typeform for launch, making a full interface will be too long. I’m going to make one for me, so I can easily process photos, but it will be done manually at the begining.
Using the “rocket launch” mode or whatever is nice but it cloud judgement, when you have the head into the code it’s hard to think, so, always good to take a step back.
Still blocking on the workflow for payment. No idea how to handle that with Stripe yet; For the launch, I’m going to make a special offer:
- People pay $5 to get 5 logos including HD versions
Time to eat dinner, will also give me a break to think and clear my mind.
Back to it!
I’m not sure about the upsell thing anymore, I think I will just offer a unique price instead with something like $15 for 10 logos, including the vectorization.
Yep, def going this option. $15 for 10 concepts and $10 for 5.
Going to have to reorganize how the back-end work but shouldn’t be that complicated.
Back-end reworked, it workd fine now!
I might actually be able to release the UI tomorrow since the new way of doing it is easier. Just need to make sure the tweaks to improve the quality are working.
Time to sleep.
Got woken up at 2AM by some drunk retards shouting outside. As much as I love Bali, it’s been turning into a carnival for drunk tourists those past months.
Anyway, time to work.
1:20 I’m struggling with the same API as yesterday. How crazy is it, their doc have EVRRY SINGLE LANGUAGE EXCEPT NODE, and the customer support is ape shit, I told them it was working in Python and they asked me if I entered my API key. Gosh, I’m loosing patience, spent 1 hour doing research and using GTP to convert the Python code that works well into node, but nope.
I just googled what I needed, found another service and it took me 1 minutes to implement it, why am I so retarded
The whole thing seems to be working. Going to add some improvements. I’m behind so I’m going to do something I never do: take a coffee.
Alright, new system + reworked user interface seems to be working.
Posted a question about market research:
Amazing way to get a better understanding on how people do their stuff without being commercial. It also generate engagement so the tweet is likely to be seen by more people.
For now, the interface is done:
And electricity just went down, nice! Means that I don’t have my second monitor for at least a few hours, gonna be slower, shit.
Now focusing on the user interface/authentification. Managed to get a no-registration form done.
Time is out, I’m now going to focus on the landing page, I’ll do everything manually to start with.
Damn, was so into it that I forgot to eat
Tbh I should have focused on making a killer landing page instead of a back-end. I thought I could handle it in 48 hours but no, too long. Anyway, I should be able to pop up somehting nice.
Landing page pretty much done. Gonna add some polish, create Typeform and I guess it should be good!
I took the most retarded name: MakeLogoAI.com
Easy to read, and domain available, I need nothing more lol
Need to do the domain redirect thing and push to Heroku. Shit, just realized I forgot to add the feedback box. I will add it after release, too bad.
Nevermind, will add crisp
Nice, took 1 minute, love it.
40 MINUTES LEFT AND IT’S NOT READY
Aannnnnnd website is live!!
Next: Add tracking and other finitions
Tracking live! Love you Simple Analytics ❤️
Just realized the video don’t autoplay on iOS, nice!
Everything seems to work, nice!!
48 hours later, to the minute, the MVP is published, yeaaah!
Going to cool down, I’m hungry and by back hurt like hell.
This feeling is amazing 😍
Made two other sales while sleeping!
Goal of today:
- Handle all the orders
- Doing groceries
I’m VERY happy with the results so far, the logos look great! Waiting for the user’s feedback to call it a win tho
Alright, I’m done. Going to do some groceries and back to work. Just saw some people starting to do the exact same thing as me, so time is against me.
A competitor with customers = Double validation.
Now that the product is made I’ll write the blog less, prob 1-3 times a day or something. Full focus on production.
Fuck that, I’m launching on Product Hunt within the next 7 days. Focus #1: Proper back-end for now.
The focus of today: Building the backend with Stripe.
Stripe is working, and the back end is pretty much done. It’s probably the cleanest code I’ve ever written. I even was able to fully set up AWS S3 using ChatGTP, so cool !!
18:25: Backend is done! Let’s make some tests to see the quality.
21:52: I’m honestly impresses by the quality and originality of the designs.
All logos were processes correctly with a few minor errors!
Today: User interface
Going to find some inspiration on Dribble
Pretty happy with the result
Looks like the checkout is done! Really happy to have a no-registration user account, everything is handled for the users. Will need to find a way to avoid entering card details every time on the long run tho.
Next step: The Logo list.
Goal of today: launching on ProductHunt and not crash.
I made a post on IndieHackers about the story, and yesterday I made a post on Twitter asking who’d be ready to support the launch so I can send them a DM. I don’t expect sales, but hopefully, I can get a top 3 badge, that would be amazing!
ProductHunt page: done!
I need to write the first comment, but this is a good thing already done. In the meantime I also hit the first page of Indie Hackers, yay!
I didn’t re-made the whole landing page, as the current one seems to work fine, so I just made a dark theme to match the app and tweaked a few things.
Holly molly, can’t believe it, I’m first on the list 🤩
After 24 intense hours, I ended up with the #1 Product of the Day 😍
I never imagined that it was going to blow up that much. I aimed for the first spot but didn’t think I’d make more than $100 in sales or something. obviously, the fact that I have some followers played a major role in there.
I’m going to focus on making MakeLogoAI THE app for everyone who can’t afford a designer. Ideally, I want people to generate quick brand kits with logos, fonts, etc. It’s going to be super challenging, but I think I can do it.
I also need to fix the mess that my personal life has become. Going “under the water” for that long totally disconnected me from reality. I live 5 mins away from the beach and haven’t been since August.
Going to slow things down, get a haircut tomorrow and . Might even go for a walk on the beach. If I have time.
Alright, woke up too late to go to the gym, so just going to work & get a haircut today.
Spent a couple of hours thanking everyone and basic customer support, now focusing on improving the app.
First thing: Improving user onboarding. Added a “requirement” field for users who have specific ideas.
Also added some social proof on the landing.
Learning to use <canvas> to allow users to add text to their logo and stuff
Result is pretty neat! I need to improve the calculations and a few other stuff, but really happy with the render, you can put any Google Font in the box and it will generate the logo.
Next: Finding a way to suggest fonts/colors using AI.
Got lots of customer support to handle due to not having email notifications ready lol, so I had a very hard time focusing yesterday.
Still managed to make the Logo Editor:
I need to fix the logo being “compressed”, but I guess that should be doable.
Now going to add the module to suggest fonts.
Published the editor as well as a few changes. Since I don’t have an email notification when the logo is ready, lots of people asked me where are their logo, so I changed the post-purchase page to have a video explaining
Time to add some tracking on the website to understand users better!
Just bought makelogo.ai! 🥳
Rest of today will be focused on tracking and stuff so I can start to make tests and measuring results;
Actually never mind, customer satisfaction is #1, so I’m going to improve that. Gonna send a survey to all my customers to ask them for feedback.
Survey results are quite terrible. Only 0% of the customers would be very disappointed if they couldn’t use the app anymore, 60% would not care, and 40% would somewhat be disapointed.
Based on the form, I’m making an improved workflow that should solve lots of issues. It’s really hard to focus with my back being completely blocked, customer service interruptions and everything else.
This is probably the lamest Christmas I had. Since I’m single since the launch day (great timing), I’m doing it alone, eating boiled chicken in my room with the AC leaking and flooding the floor. But hey, at least I have a hot shower and food.
I totally remade the back end from scratch, this new version should be way better and more scalable. Spent the last 3 days on it, I can’t wait to start marketing.
I have trouble focusing and working, I’m really tired those last few days. I guess the fact that I didn’t take a day off since August plays a part. Need to push things a bit further and then I’ll take a day or two off.
Discipline > Motivation
Had to really drag me on the last parts, but the app is now done. Some features will be added later, but it’s not big deal, most of them are already there. Was very hard to be productive, I really need to take a small break I think.
Now that the app is done, I can focus on the “technical marketing”, which includes tracking, email automation, etc.
Gonna start by making an event-tracking system that can easily be plugged in with external software.
Basic event system is done, now going to focus on plugging email marketing.
Bento works pretty well, not perfect yet but good enough for my needs atm! Next, Google Analytics to add some additional tracking + split-test
Wow, this is actually way more complicated that it should. Seems like none of the tools have an easy way to track purchases and other events easily. That’s so dumb!!
Finally found a way to get the order value, now I just need to plug it to Facebook and Google analytics
Seems like it worked. Need to wait until data flow in to verify. Next step, sharing cookies between the domain and subdomain.
Last few days have been slow, but the data seems to be working.
Made a dashboard to check how is the website performances every day.
Fixed a few bugs, now focusing on adding some SEO to the app.
Also need to improve my tracking system, as apparently not all events were matched properly.
Spent most of the day fixing a major bug, had to re-write a part of the app. Hopefully it won’t crash during the night.
Seems like the new version didn’t crash, yay!
Started to run tests with Google Optimize to improve the landing page. Testing small things to get familiar with it first, then going to test different layouts.
While data are gathered, I’m going to implement a user feedback loop to allow me to measure user satisfaction. So far I tried to send a typeform, but only 8 people answered on the hundred of orders I had, so, doesn’t seems to work.
Instead, I’m going to add something directly inside the app.
Seems like Google Optimize is working well. Going to test landing page variants.
So far there is almost as much traffic from mobile than desktop, so I can’t prioritize one over the other, even tho Purchases happen on Desktop in 70 of the cases (which means mobile needs improvement)
Made a few more tests, and added an affiliate program so people can get free logos if they refer people. Will add $ rewards later.
The #1 issue is customer service. I have doing it, and it really break my focus, so I’ll hire an assistant to help me out with that. Hopefully it will allow me to focus on improving the app.
I just sold the app for $65,000