The Ex-Optician Who Learned to Code From a 15-Hour YouTube Course — How a “Never Reinvent It, Just Make It 1% Better” Clone Strategy Took StoryShort to $20K/mo (and a $35K/mo Portfolio)

StoryShort turns a blog URL or a one-line topic into a finished ‘faceless’ short: the AI writes the script, adds a voiceover, composites captions and visuals, and outputs videos for TikTok, YouTube Shorts, and Instagram Reels — even auto-uploading on a schedule. GPT-class models handle the script; a voice AI handles narration.

It was built by Samuel Rondot in France. He used to be an optician and couldn’t write a line of code a few years ago. He quit his job, taught himself from a 15-hour YouTube programming course, and applied each lesson directly to his own product — that’s the heart of the story. Today he runs three SaaS apps: useArtemis (a LinkedIn sales tool, $15K/mo, 10,000 customers), StoryShort ($20K/mo, 4,000 customers), and Capacity.so (an AI coding helper, still early at ~$900/mo) — together ~$28K–$36K/mo, all bootstrapped with zero outside funding.

StoryShort is the breadwinner. It launched in August 2024 and ramped so fast that by that November he was posting on X that ‘my newly launched SaaS is earning as much as my 2-year-old one.’

What actually drives it isn’t a viral moment — it’s how he *chooses*. Samuel’s single rule: never reinvent the wheel — pick something that already exists and shows signs of working, then make it 1% better. He filters candidates by ‘would I use it myself / does it already work / is there clear traction,’ moving the bet away from novelty and onto execution speed plus a small improvement. For him, AI wasn’t magic; it was the tool to rebuild validated demand quickly out of modern parts (GPT, voice AI).

From the founder (primary source)

Samuel Rondot

@samuelrdt

·Follow

My newly launched Saas StoryShort.ai is now earning as much $$$ as my 2 yo old one useArtemis.co My thoughts 👇 - B2B is hard and and having companies as customers do not reduce churn. - I spent way too much time on useArtemis and should have give up way Show more

8:47 PM · Nov 8, 2024

51

Reply

Copy link

Read 15 replies

The repeatable playbook

1. 1Don’t invent from scratch — pick only markets that already exist and show traction
2. 2Filter ideas by ‘would I use it / does it already work / is there clear traction’
3. 3Differentiation needn’t be world-first: aim for ‘1% better’ (faster / cheaper / simpler)
4. 4Don’t train a giant model; translate modern AI (GPT, voice AI) into the product to rebuild fast
5. 5Prefer thousands of B2C subscriptions over a few B2B contracts (often easier to acquire and retain)
6. 6Don’t rely on virality; win SEO surface area with lots of on-theme content
7. 7Build in public on X (share numbers and lessons); don’t bet on one app — run a portfolio

Deep dive

【Deep dive】We break down, in order, how a non-engineer reached three SaaS apps and ~$35K/mo. The crux is two choices — *what to build* and *who to sell to* — not raw technical skill.

■ Clone only validated demand, then make it 1% better. Samuel’s decisions start from ‘don’t invent from scratch.’ StoryShort itself entered an already-selling market — automated faceless video — and rebuilt it. The lesson for indie builders is heavy: chasing a brand-new category is a gamble, but picking a market where money already flows lets you skip the hardest step, demand validation. Differentiation doesn’t require ‘world first’; a little easier / faster / cheaper than the incumbent is enough. He filters candidates by ‘would I use it, does it already work, is there clear traction,’ shifting the bet from novelty to execution speed.

■ B2C beats B2B on churn — a lesson from experience. His first hit was the 2-year-old useArtemis (a B2B sales tool), yet he says plainly on X that ‘B2B is hard, and having companies as customers does not reduce churn,’ and that he ‘spent way too much time on useArtemis.’ The later StoryShort — B2C for individual creators — matched and then passed useArtemis’s revenue in far less time. The implication is clear: for a solo founder, stacking thousands of individual subscriptions can be easier to acquire and retain than chasing a handful of enterprise contracts. The intuition that ‘big B2B = safe’ isn’t always right.

■ Diversify with a portfolio — don’t bet on one. He didn’t pour everything into his one hit; he runs useArtemis, StoryShort, and Capacity.so in parallel. If one plateaus or turns into a red ocean, another can carry the total. Clone strategies tend to raise the odds of a hit while capping the ceiling, which pairs well with holding several. A solo builder can reach full-time income by stacking small, reliable SaaS like building blocks — a repeatable model.

■ Growth is SEO and build-in-public, not virality. Where many apps spike on a single TikTok, StoryShort’s growth is unglamorous and cumulative. It builds a large body of content/blog posts aligned to its theme (video and content creation) to win SEO surface area — search traffic being a slow-decaying asset — while Samuel keeps doing build-in-public on X (sharing numbers and lessons) to earn discovery and trust in the solopreneur scene. Less flashy, but for solo operation it’s more durable than virality that dries up the moment ads stop.

■ Use AI to *rebuild fast*, not to ‘build.’ Technically, StoryShort generates scripts with GPT-class models and narration with a voice AI — translating existing powerful AI into a product experience rather than training a giant model. A former non-engineer got this far because AI lowered the implementation wall, letting him reassemble validated ideas out of modern parts quickly. Once the barrier drops, the contest moves — from the start — to *what to build* and *how to distribute it*. That’s the central lesson of this case.