• signal5
  • Posts
  • ๐Ÿš€ How Aragon AI Grew to $10M Revenue (ARR)

๐Ÿš€ How Aragon AI Grew to $10M Revenue (ARR)

โ€œBecoming is better than beingโ€ โ€• Carol S. Dweck

Aragon.ai reached $10 million ARR in just under 2 years with a 9 person team

At 25, Wesley Tian was a college student disillusioned with academia when he stumbled upon a research paper that would change everything. With just a small team of 9 engineers, no sales team, and limited venture funding, they built Aragon.ai into a $10M ARR powerhouse.

Their secret wasn't fancy offices or bloated teamsโ€”it was lightning-fast execution, relentless experimentation with affiliate marketing, and a laser focus on solving a simple problem that affected millions: Why should anyone pay $250 and waste half a day for professional headshots when AI can do it better in minutes?

The TL;DR: From an AI research interest to $10.8M annual revenue in just 3 years by solving a simple problem - expensive, time-consuming professional headshots!

๐Ÿ’ก Problem Solved:

  1. Professional headshots cost $250+ and require booking, travel, wardrobe prep, and hours of time

  2. Aragon created an AI solution that turns selfies into studio-quality headshots in minutes

  3. Now serves 750,000+ users who've generated 25M+ headshots

๐Ÿ› ๏ธ How It Works:

Upload 6+ selfies โ†’ select outfit/background styles โ†’ AI generates 100 professional headshots in minutes

๐Ÿ”„ The Pivot Story:

Wesley Tian was exploring tech ideas when he discovered Stable Diffusion and Dreambooth. After experimenting, he built an AI avatar product and noticed users loved the professional headshot style, so he doubled down on that specific use case. Indie Hackers

๐Ÿ“ˆ Growth Strategy:

  1. Started Simple: Built prototype in just one day using Gradio

  2. Guerrilla Marketing: Posted strategic messages on Reddit, Blind, and other platforms

  3. Affiliate Program: Launched in Spring 2023, affiliates quickly generated over 50% of revenue, scaling to almost 2,000 affiliates within two years Rewardful

  4. Content Marketing: Built audience on Twitter/X and LinkedIn through authentic posts

๐Ÿ’ฐ Revenue Journey:

From launch โ†’ $3K first month โ†’ $900K monthly revenue โ†’ $10.8M+ annually in 2025

๐Ÿ’ช Competitive Edge:

Backed by AI researchers from Meta and Microsoft, Aragon can generate realistic headshots with minimal input photos at lightning speed compared to competitors requiring more uploads or longer processing times. AI Mojo

๐ŸŽฏ Customer Focus:

  1. Job seekers needing professional LinkedIn/resume photos

  2. Businesses wanting consistent team headshots

  3. Sales professionals improving conversion rates

  4. Therapists, physicians, and gig workers boosting client trust

๐Ÿ”ฎ What's Next:

Expanding beyond headshots to become "the next era of photography" with AI solutions for wedding photos, dating profiles, healthcare professionals, and custom image editing tools Aragon

๐Ÿ”‘ The Winning Formula:

"Find a simple, expensive, time-consuming problem that AI can solve faster, cheaper, and at scale - then build distribution mechanisms that reach your target audience where they already are."

๐Ÿ‘‡ Your Turn:

What other expensive, time-consuming photography needs could AI solve next? Have you tried AI headshots? Try it here

Current holdings - for the quarter

If you had to own one stock today - we would say own NVDA

If you had to own 3 stocks today - we would say own - NVDA, COST, AVGO

If you had to own 5 stocks today - we would say own - NVDA, COST, AVGO, META and LLY

Not investment advice

โ€œWe like to think of our champions and idols as superheroes who were born different from us. We donโ€™t like to think of them as relatively ordinary people who made themselves extraordinary.โ€

Carol S. Dweck

PS: If you want to read more of Ms. Dweck - Read this

How well are you aging? Test yourself

Amazon will reveal more headphones for the gym than it seems there are actual gyms. Finding the right secure-fitting, water-resistant, lightweight wireless earbuds is not easy.

We have found exactly what you need