A Complete Breakdown To How We Built A $20K MRR AI App
Hello! Who are you and what business did you start?
I am Ricardo Ghekiere, jack of all trades, but master of none. I dropped out of school when I was 20 to start a coffee bar, probably the most unsexy business you could start. 1 year later, we reached 3 locations and were forced to sell my shares which made me lose all the money I had invested and returned back to finish business school.
From there I sold cocktail boxes, sold a marketing agency, went back to running a marketing agency (saasmic.io) and on the 1st of December 2023 decided to partner with Miguel Rasero (CTO and Co-founder) to grow Betterpic.io after a drunk night in Spain.
What started as conversations about a potential acquisition, turned into a lasting partnership with a clear vision on how we want to democratize professional imagery. Helping busy professionals and job seekers turn their selfies into professional headshots under 60 min starting from 29$.
When I acquired the business, BetterPic was averaging around 1500$ per month. In July 2024 we closed the month with +/- 20k$ of gross revenue, increased our team to 5 people (+2 founders) and generated over 11M+ headshots.
What's your backstory and how did you come up with the idea?
Hey there! Miguel Rasero here, the other co-founder behind BetterPic. Let me tell you back to where it all started.
I've got this never-settling, curious mind. When something grabs my attention, I can't stop until I've figured it out. Software development, nutrition, business strategy, economics, AI - I've obsessed over them all. And I always need to apply what I learn. That's how this whole story begins.
Picture this: It's 2014, I'm 14, armed with some basic Java knowledge and a Minecraft addiction. One day, scrolling through the chat on a Minecraft server, I see: "This server is on sale, contact xxx for details." My mind starts racing - what if I could create the minigames I'd love to play? So, I reached out to the owner.
Turns out, the server was a mess - lag issues, angry users, barely breaking even. The owner wanted out, offering it for a steal (about 500€, if memory serves). That looked like a cool challenge. With my whole life savings (245€) and a little loan (255€) from dad, I dove in headfirst.
Over the next four years, I transformed that struggling server into Lutorlandia, one of the top Spanish-speaking Minecraft servers. We served over 2 million players and got hands-on experience in everything - DevOps, software development, marketing, team management (we peaked at +25 people), cybersecurity (we even dealt with a DDoS attack big enough to take down a small city), finances, and even dipped into machine learning for our anti-cheat system.
Post-Minecraft, I jumped into various projects. I managed the entire Azure infrastructure for JovidApp, a job search application that I co founded. Then came an e-commerce venture in 2019, teaching me about digital marketing and dealing with the Asian market. From 2020 to 2023, I helped small businesses boost their digital presence, improving my skills in SEO, web development, and digital advertising.
Fast forward to my end-of-university project. I wanted to make something real, not just churn out an 80-page document. I'd been playing with GPT-3, and I knew - it had to be about AI.
Then, boom! February 27, 2023 - I tested a tool generating fantasy-style selfies. The results blew my mind.
That's when it hit me: create the best solution for professional profile pictures. Google Trends showed growing demand.
I had my topic. BetterPic was born. Ready to dive in another rabbit hole.
The learning curve was steep, but I was ready. With my background in statistics, data analysis, and programming (Python and especially JavaScript were my go-to languages), I dove into open-source models like Stable Diffusion. The initial results were... interesting. And by interesting, I mean BAD. But that's the beauty of working with emerging tech - you're right there on the frontier.
Enter my first guinea pig: my grandpa. Now, picture a traditional spanish cordobés - yep, hat and all. Hat in every. Single. Picture.
But, even with the not-so-great quality, his reaction? Pure gold. Surprise, delight. Bingo! We had our wow factor (or at least, we had my grandpa's wow factor 😅).
Three months of AI crash course followed. Text-to-image models, input picture processing, pipeline optimizations, advanced training methods, lots of postprocessing and testing with custom models, you name it. By June, I had something decent. Facial resemblance? Check. Weird textures? Mostly gone. I even convinced grandpa to use it for his Facebook profile. First customer, check. Paying customer? Well... I had to start somewhere!
Oh, and if you're wondering if people noticed it was AI - surprisingly, nada. Zero.
Quick side note for my fellow perfectionists out there: GET. YOUR. STUFF. OUT. THERE. Seriously. Get feedback, especially the bad kind. It's gold. Tell people, "Don't tell me what you liked, tell me what you DIDN'T like." Otherwise, especially if they're your pals, they'll sugarcoat it.
And if you're launching a paid product? Value those paying customers' opinions. Free product feedback can be deceiving - they might not value it the same if they were paying.
To sum it up, let me borrow some wisdom from Aaron Levie, Box founder: "Do something that wasn't possible 3 years ago. If you could have done it 5 or 10 years ago, it might be too late, or it might not be different enough to make a lot of people care, and that's ultimately what it takes to scale."
So, let's check our boxes:
- Market being created (blue ocean, but sharks incoming - got to move fast)? Check.
- Literally impossible months ago? Double-check.
- Initial model and pipeline ready? You bet.
- Initial wow effect? Some jaws dropped.
- Use case validation? Early signs!
Time to build an MVP folks.
Take us through the process of building the first version of your product.
Launching BetterPic? Imagine conducting an orchestra... with just one musician. Me. As a solo founder, I wore all the hats: marketing guru, legal eagle, tech wizard, customer support champion, financial planner, and..."designer" (spoiler: early designs were more 'functional' than 'beautiful').
The AI model was nearly ready, but getting it online was the real party. I dove into Vue 3 (Nuxt) as frontend, Supabase for logins, PostgreSQL for the database, Vercel for deployment, and a simple Node.js backend. Voila!
Oh, and let's talk about that logo. Picture this: me, bleary-eyed at 2 AM, armed with nothing but Canva and sheer determination. Three hours later, boom! Logo done. Was it good? Probably not, I’ll let you decide. Served its purpose? Sure it did!
Here's how the website looked back then:
SEO? Started that before the website was even decent. Months of churning out blog posts (thanks, ChatGPT) and building links. Planting seeds for a garden I hadn't designed yet.
August hits, and I get this crazy idea: launch before vacation. Some crucial parts weren't ready (like, you know, showing people their AI generated headshots), but who needs a complete product, right?
Day two of vacation. I'm lounging on the beach, feeling great, finally resting, relaxing, nice. Then my phone pings. Congratulations! You've got your first customer!
Oh, sh*t.
Imagine me, on the beach, laptop balanced on my knees, coding like a caffeinated squirrel. Sunscreen and sweat dripping onto my keyboard, trying to build half a product on the fly.
I fired off an email to the client: "So sorry for the delay, we're absolutely swamped with orders!" Translation: You're our one and only customer, and I'm sweating bullets here, trying to deliver what I promised while fighting off seagulls and praying my laptop doesn't overheat.
Somehow, someway, I managed to pull it all together in a day. Fixed bugs, connected APIs, probably invented a few new coding techniques fueled by panic and piña coladas. Still not sure how I did it - pretty sure it's impossible to replicate without the specific combination of sunburn, stress, and sand in my shorts.
But you know what? That customer got their AI pics, loved them, and probably had no idea they were made by a guy who looked like a lobster with a laptop. And me? I learned that sometimes, the best problem-solving happens under pressure - or under the sun, in this case.
Lesson learned: maybe don't launch right before vacation. Or do, if you enjoy coding with a view and a side of adrenaline. Your call!
And then, setting up the company in Spain? What a fun time. Bank meetings, notary visits, waiting... so much waiting. I couldn't even find a fitting category for "AI image generation". Seriously, Spain, can we add at least "cool AI stuff" to the list of possible businesses?
Traditional banking was so slow, I had to use Revolut just to start taking money through Stripe. Sometimes, you gotta dance around the system to keep things moving.
Despite the chaos, BetterPic made some sales (tbh, more than expected):
- August 2023: 343.24€
- September 2023: 2,448.84€
The start of BetterPic? Pure insanity. The MVP wasn't pretty (seriously, the design was BAD), but it worked. Even if the site felt like a total scam, people were using it, surprisingly.
Every little win felt huge. Every problem was a chance to learn. From beach coding to legal mazes, and drinking way too much coffee, it was a fun experience. And I wouldn't change a thing. Well, maybe the design. Definitely the design.
Describe the process of launching the business.
Challenge 1: Low Conversion Rate on Website
One of the challenges faced was the feedback we received on our website:
“It looks like you are going to scam me” or “It looks like a school project” weren’t uncommon things we heard when showing the website to people.
I came to a point where I noticed myself talking about it but mentioning that we are currently rebranding it before people would even search for the website.
I’ll let you be the judge of it as this is how it looked like on december, few months after the MVP:
This is also reflected in our sales figures when we looked at our website visit-to-purchase ratio which we have set up in Mixpanel.
Yup, only 0,63% of our website visitors convert into sales.
These metrics would never be able to get us into profitability, especially given that most of our users are still taking the lowest package of 25$
We decided to do something about it and spend an entire month rebranding every single screen you could find.
What started with “let’s just change a few colors" ended up taking a full month to rebrand every single screen we had.
As our CTO, Miguel mentioned in a conversation: ‘If we killed the legs, arms, and the body, can we still call it a rebranding or is it just a new company’?
Here is a quick overview how the rebranded B2B platform for company headshots looks like.
Challenge 2: Increase our Average Order Value through upsells & pricing changes
Running growth is not just about changing your bidding model on Google or adapting your targeting strategy on Facebook.
Across the years, I have learned that most of your impact is not improving marginal gains in the channels but affecting the product, which impacts the channels you distribute it on.
One of the product questions we started asking ourselves is: ‘Why are over 50% of our users buying the cheapest plan (25$)?’.
Here is a quick overview of our package pricing:
This is a pain point I highlighted during a presentation I gave on an entrepreneurship talk last month:
As you can see from the purple circle (in our mixpanel account), the largest amount of packages we are selling is our 25$ package, which, according to my ‘awesome financial plan’, would never bring us into profitability. Pretty important as a bootstrapped business if you ask me.
So, we decided to run an experiment with the following hypothesis.
Hypothesis:
Most people choose our $25 plan over our $35 plan because we believe the difference of $10$ is too big. As most people don’t fully understand the differences between packages, they rather go with the cheapest package to ‘test it out’.
Experiment:
Increasing the price of our basic from 25$ to 29$ without changing anything within the package will make the 35$ package seem more reasonable and, therefore, increase our Average Order Value.
So here is what happened.
Our middle package (35$) saw a 400% sales increase, making it our best-selling package, even more so than the 25$ or, in this case, 29$ package.
Increasing our average order value to drive our company to profitability.
But you might think, what about our conversion ratio from visit to purchase?
Luckily, that has increased monthly, surpassing a 5% conversion ratio in the USA (our key market).
Since launch, what has worked to attract and retain customers?
With the low average order value (around 35$) , we had to decide to get into the organic marketing (SEO) game early on to drive profitable growth in the long term, and Apoorv from Saasmic has been leading the team in increasing our organic keywords month over month. Here is a quick snapshot from our Semrush account showing our ranking for 1,4k keywords (starting from 537 when we started).
We are pushing our position from important keywords like ‘studio AI headshots’ from position 43 to the first page of Google (8). We all know that the second page of Google is where SEO people hide bodies.
Need some headshots in Bordeaux or München? No worries, you’ll find us in the number 1 spot to get your headshots.
All efforts are slowly helping us to increase our monthly organic revenue, and this is while we are still catching up at lightning speed.
Yet not all efforts went very smoothly, as with everything in life when you experiment. As you can see from the screenshot below, we were ranking for an important keyword, ‘Ai headshot generator’, in April.
But then Apoorv decided to make some changes to our website to better align our content efforts towards this keyword, which ended up in almost drama as we lost our ranking in May for one of our critical keywords.
We decided as a team to trust the process and not adapt anything, which is also when, in June 2024, we suddenly pushed our position from nowhere to find to the 25th position. Not 1st place yet, but it's a good direction already.
What is Apoorv's goal for SEO?
Hit 10k in organic revenue by the end of the year by using a smart combination of blog posts, programmatic SEO through our near me pages and cool free tools we are giving away.
For example, we recently launched our LinkedIn AI Picture Analyzer Tool, allowing users to upload their current profile pictures and get a detailed overview of how to improve them.
That’s not all on the organic side.
I believe in stacking your odds, where multiple channels support each other for your growth story.
This is why we’ve been driving organic visits from Pinterest through our automated system, which we started early on in our journey to publish over 100+ headshot pictures of men and women on Pinterest.
So far, we have driven over 750k+ impressions on our Pinterest page, allowing people to discover us on a monthly basis without paying a cent.
Allowing our pictures to be saved and lead to our website to drive sales.
Funny enough, none of us has ever done Pinterest marketing. We just decided to trust our gut, exploit a system, and deploy our marketing knowledge. Sometimes that’s all you need.
Other than that we work closely with influencers, run ppc ads on google and reddit as we focus on profitability first as we are fully bootstrapped and the majority of our cash goes into finding great people to build the product.
How are you doing today and what does the future look like?
So, let’s start with what we call our Monthly Performance Bingo of July 2024:
The Good:
- Gross revenue: 20,016$ (+90%) vs 10,555$ LM --> almost doubled revenue in 1 month :melting_face:
- Profitability: +3680$ --> We pushed our first month into profitability (excluding salaries) :fire:
Product:
- NPS: Pushed to 3,11 (from 2,9 LM) --> still a long way to go but heading in the right direction.
- Refund rate: 6,78% (vs 7,69% LM) --> still not our 5% goal yet getting closer.
- Studio Upsells: 5% vs 3% of users LM now buy extra credits in our Studio.
In short, we are closing of the month with some great figures where we 10x the business in 6 months time
But it's better shown when you look at the growth month over month from the beginning when Miguel Rasero and myself partnered up:
We started with a website conversion ratio of 0,68% worldwide and pushed all the way to 1,8% right now, with our key market up all the way to 5%.
Our target by the end of the year is to hit 50K in monthly revenue, followed by a capital raise beginning next year when we finalized the Founders Institute Accelerator who is preparing the company for the best possible structure for growth.
Through starting the business, have you learned anything particularly helpful or advantageous?
Getting AI headshots is extremely hard, there are so many factors you have to get right and we are never finished tweaking and improving our model. One interesting psychological phenomenon happens when people evaluate their own professional AI headshots.
There is a fun psychological moment that happens on how you see yourself vs how other people see yourself. Most people look at their pictures and like the ‘improved version’ of themselves as that is what they are looking for in a professional headshot, yet when you share the picture with somebody who is close to them they would argue that the liked pictures aren’t them but some other pictures are as that person is looking for a picture that resembles the closest to how they actually look. This psychological phenomenon can be described as self-perception versus external perception.
What platform/tools do you use for your business?
I am a big fan of customer.io to create marketing automation flows as we are in a high volume business and one-to-one communication isn’t always feasible, yet we do love to be proactive. I pretty much try to automate everything I can in the business as I am pretty lazy from nature and due to my ADHD I hate repetitive tasks.
My favorite flow is when people leave an NPS score they will be funneled into a workflow where depending on their score they will receive a personalized message from myself to further explain the score to drive the most insights on how to improve the product and automatically create a ticket with support to make sure we answer every single person.
I also use Mixpanel on a weekly basis to understand how the business is doing, run reports or see which traffic is converting the most.
What have been the most influential books, podcasts, or other resources?
Acquired Podcast is one of the best podcasts to understand how companies started out and the underlying business model.
I read the book ‘good to great’ by Jim Collins when I was 19 years old. I didn’t fully grasp it but knew the book would be valuable at one point. It was only 5 years later when building a team I fully understood what the book was really about and was always left with the words ‘get the right people on the bus, and only then figure out together where to drive towards’. This is exactly what we are doing at Betterpic, we hire talent for them to tell us what to do, not the other way around.
Advice for other entrepreneurs who want to get started or are just starting out?
The power of installing feedback loops in your company
When I was 21 I launched a company called Alchemistbox, a way for people to create 4 perfect cocktails at home without the hassle of finding the perfect recipe or buying expensive bottles they won't use again.
In fact, the company was launched when I was finishing up my last year of business school as I was too lazy to write a thesis and convinced the professors that launching a real business was more effective. You can find the worst pitch deck ever created by me right here.[
](https://www.slideshare.net/ricardoghekiere/alchemistbox-pitch-series-a-64852578?utmsource=blog.saasmic.io%26utmmedium=referral%26utm_campaign=how-feedback-loops-helped-me-sell-my-company)This is what it looked like as a final product when we started scaling (yeah yeah, I should have had a good haircut before the news came over).
As you have probably guessed, this is not how the product started out
In fact, this is the first time I am sharing what the first prototypes looked like when starting out as I had to dig deep into my picture library.
Because as the famous Reid Hoffman's says — “If you are not embarrassed by the first version of your product, you've launched too late.".
This is what the first prototype looked like:
Yup, it was just an ordinary shoe box, bottles from the pharmacy across my door, and a poorly designed 'how-to-flyer' for which I paid a student exactly 40€ for.
Don't even get me started on the 'label' design.
The product was far from ideal, but I had learned an important lesson from my first failed start-up where I believed IKEA was going to steal my idea.
Get out to the market as soon as you can to gather feedback from customers. It's the only way to win.
It was time to get out of my house and gather my first feedback moments.
The First Sale of Alchemistbox using a Feedback Loop
The product was 'ready'. Now it was time for sales to flow in (as good founders always believe it will rain €€€).
So I decided to grab a booth at a pop-up concept area in Antwerp, Belgium.
Pitching Time.
Every morning from 8 am until 6 pm I would be pitching the product.
I felt like an old-school paper boy shouting on the corner of the street.
"The perfect way to create 4 cocktails without finding a recipe".
"Create cocktails at home without buying big bottles".
As you can imagine sales exploded and sold out on the first day.
Just kidding.
in 4 days, we sold 0.
Like not even 1 freaking box.
Literally 0.
Something was wrong, and I just couldn't figure it out.
Until the last day of the fair. Something magical happened.
An older couple walked up to one of my boxes, pointing at the exact picture that I had pasted into my box.
"Oh, this guy on the left really looks like Steve. We should buy it as a gift for him".
I was shocked.
My usual pitch was "Buy it for yourself to create the perfect cocktails".
And now this older couple just decided to change my whole idea and buy it for somebody else.
But guess what? They took out their wallet and paid 45€ for my first box.
The first sale was made.
That last day, I adapted my pitch completely based on the feedback I received when making my first sale and was able to push another 9 sales by just tweaking the exact same product towards a different use case.
My best tip when building a company?
Install as many product feedback loops to drive insights about your target audience.
The faster the feedback hits, the quicker you know what to create to drive further growth.
Are you looking to hire for certain positions right now?
Yes, we are looking for a couple of roles:
- Content Strategist
- AI Engineer Intern
More info on the roles.
Where can we go to learn more?
If you have any questions or comments, drop a comment below!
Download the report and join our email newsletter packed with business ideas and money-making opportunities, backed by real-life case studies.
Download the report and join our email newsletter packed with business ideas and money-making opportunities, backed by real-life case studies.
Download the report and join our email newsletter packed with business ideas and money-making opportunities, backed by real-life case studies.
Download the report and join our email newsletter packed with business ideas and money-making opportunities, backed by real-life case studies.
Download the report and join our email newsletter packed with business ideas and money-making opportunities, backed by real-life case studies.
Download the report and join our email newsletter packed with business ideas and money-making opportunities, backed by real-life case studies.
Download the report and join our email newsletter packed with business ideas and money-making opportunities, backed by real-life case studies.
Download the report and join our email newsletter packed with business ideas and money-making opportunities, backed by real-life case studies.
