Basics
| Item | Details |
|---|---|
| Founder | Danny Postma (Dutch, based in Bali, @dannypostmaa) |
| Product | HeadshotPro — AI professional / business headshots |
| Team size | 1 person (zero employees) |
| Revenue | $300K/month, $3.6M ARR |
| Pricing | From $29 (one-time, not subscription) |
| AI stack | Stable Diffusion + DreamBooth fine-tuning (self-built pipeline) |
| Prior exit | Headlime (AI copywriting), sold for $1M at $20K MRR (only 8 months after launch) |
| Source | Starter Story / SupaBird deep dive / Rewardful case study |
Storyline
Background: Danny Postma is a Dutch indie hacker living in Bali. Before HeadshotPro, he had already exited one AI product — Headlime (AI copywriting), sold for $1M eight months after launch at $20K MRR. That exit gave him two key lessons: AI products can ramp up extremely fast, and “selling shovels” (B2B tools) is steadier than B2C.
Turning point: In March 2023, Danny noticed AI portrait generation (Stable Diffusion + DreamBooth) had matured to commercial quality, but no vertical product existed for “professional business headshots.” The key insight: LinkedIn users need professional headshots, but a traditional shoot costs $200–$500 plus scheduling. A $29 AI headshot is a perfect price substitute. He didn’t ship a product and then look for users — he validated demand with SEO first: search volume proved that “AI headshot” had plenty of organic traffic.
Outcome: Revenue scaled quickly to $300K/month after launch. Danny’s core strategy isn’t a tech lead (the AI portrait space is crowded) — it’s distribution lead. SEO organic traffic + affiliate promotion ($50K+/month, 15%+ of revenue) compound into a hard-to-copy growth flywheel. One person, zero employees, running a $3.6M ARR business remotely from Bali.
Workflow Breakdown
Minimum loop analysis
| Step | What he did | AI involvement | Notes |
|---|---|---|---|
| 1. Think it through | Noticed AI portrait tech was mature + LinkedIn headshot demand unmet; validated with SEO volume | Low | Direction call is human, but data validation is fast |
| 2. Write a spec | Minimal product definition: upload photo → AI generates → download. One-time $29 | Low | Danny’s style is “ship the smallest version first” — no detailed spec |
| 3. Implement | Self-built Stable Diffusion + DreamBooth fine-tuning pipeline | Medium (~60%) | The core AI pipeline needs manual quality tuning — not one-click |
| 4. Validate | Paying users validate quality — the $29 price lowers the trial barrier | Low | A paying user is the validation |
| 5. Iterate | Continuously improve generation quality + SEO + affiliates | Medium | SEO content may use AI assist, but strategy is human-driven |
Tool stack
- Stable Diffusion + DreamBooth: the core AI engine. DreamBooth fine-tuning teaches the model the user’s face and generates personalized business headshots. This is the technical core, but Danny hasn’t publicly shared exact model versions or training parameters.
- SEO stack (specific tools not disclosed): Danny’s core growth engine isn’t AI, it’s SEO. He systematically optimizes search rankings so “AI headshot” keywords drive organic traffic.
- Rewardful (affiliate management platform): runs the affiliate program; $50K+/month comes from affiliate channels — a perfect application of “the time AI frees up should be spent on distribution.”
Roles one person carried
| Role | AI replacement | Notes |
|---|---|---|
| Product / tech | 60% | The AI generation core is automated, but quality tuning is manual |
| Marketing / SEO | 30% | Strategy and content frameworks are human; AI may assist content production |
| Customer support | 80% | “Product-as-a-Service” model + auto-refunds minimize support load |
| Finance | 90% | Stripe one-time payments — fully automated |
| Operations | 70% | The affiliate program is automated through Rewardful |
3 Reproducible Key Moves
-
Validate demand before building: Danny didn’t ship and then hunt for users — he confirmed demand for “AI headshot” with SEO volume first. How to do it: check keyword volume on Google Trends / Ahrefs. Daily searches > 1000 with low competition = green light.
-
One-time payments > subscriptions to lower decision friction: $29 one-time has much higher conversion than $20/month subscription. For a low-frequency need like “take one business headshot,” one-time payment is the right model. How to do it: analyze your product’s usage frequency — one-time for low frequency, subscription for high. An auto-refund mechanism (refund if dissatisfied) further lowers user anxiety.
-
Replace a sales team with an affiliate program: $50K+/month from affiliates is the equivalent of “hiring a sales team paid only on closed deals.” It’s the highest-leverage growth move for a one-person company. How to do it: set up an affiliate backend with Rewardful / FirstPromoter, offer 15–30% revenue share, and let promoters write reviews and content for you.
Limits and risks
- The AI portrait space is brutally competitive: Aragon AI ($10.8M ARR), Profile Picture AI, and many more. Danny’s moat isn’t technology (AI portrait generation has gotten much easier) — it’s SEO ranking and the affiliate network. SEO rankings can be overtaken.
- Opaque tech stack: Danny hasn’t publicly shared the specific AI models and training process, which lowers reproducibility.
- The LTV problem of one-time payments: $29 one-time means you have to keep acquiring new users — no subscription compounding. If SEO traffic drops, revenue decays fast.
- Single-product dependency: Unlike Pieter Levels’ portfolio strategy, Danny’s revenue is highly concentrated in HeadshotPro.