CreatorHouse

Hey Fazier! I'm Salah, the solo founder of CreatorHouse. I built it after watching creators spend hours scrolling competitors, typing out hooks by hand and guessing why a Reel took off. CreatorHouse tracks the accounts in your niche, flags the videos that beat each creator's usual reach, breaks down the hook and structure, and turns one into a script about your own topic, in your voice. The Instagram and TikTok transcript tools are free with no signup, so you can try the breakdown part right away. I'd love your feedback, especially on what you'd want to see after you find a winning video.

The outlier score relative to each creator's median is a smart way to find real hits instead of just big accounts. How many recent videos does the median use, and does it work for small accounts with only a few dozen posts?

nice to turn the viral reels to to script

Scoring each video against the creator's own median is a smart way to cut out big-account noise. One thing I'd find useful: when a format goes viral, dozens of accounts post the same template within a few days. Could CreatorHouse group videos by sound or format across the accounts you track, so you can spot a trend spreading, not just one creator's outlier?

lean idea. Finding outlier Reels in a niche usually takes hours of scrolling.

The outlier score (views relative to a creator's own median) is a smarter signal than raw view counts for finding what really worked. How does it handle smaller niche accounts with only a handful of videos, where the median can be noisy?

Great to see really good pricing and features

The outlier score is the right metric — view counts alone don't show why a video beat that creator's usual reach. Since you asked for feedback: a way to save and tag winning hooks by niche would be huge for spotting patterns over time.

Repurposing top-performing Reel hooks into personalized creator scripts is huge for scaling short-form content output without sounding generic. Does the prompt engine adapt to specific creator tone presets, or does it learn directly from past uploaded videos?

Good job, Salah. Good luck.

Turning winning Reels into scripts in the creator's own voice is the right wedge - generic AI scripts are why most creator tools get abandoned after a week. The hard part is voice modeling: it only takes a few off-tone lines for the creator to stop trusting the output. Curious how you're handling that - few-shot examples from their past Reels, or a style profile they can correct? A quick thumbs-up/down on each generated script would also give you the training signal most tools in this space never collect.

Love the focus on "outlier score" rather than vanity metrics—makes competitive research significantly more actionable without endless manual scrolling. The free transcript tool with no signup is also a great touch to lower friction. Congrats on the launch, Salah!

Scoring outliers against each creator's own median is a smart choice: it separates a strong hook from an account that is simply large. Does the remix keep the original pacing (hook length, cut rhythm), or only the script structure? It would help to see which elements of the source video the new script deliberately keeps.

It work for any niche ?

Congrats on the launch! The outlier score is a clever angle — comparing a video to the creator's own median is much more useful than raw views. Turning a proven hook and structure into a script in your own voice should save creators a lot of research time.

Turning winning videos in your niche into scripts in your own voice is a smart angle. How do you capture someone's "voice": from their past content, or from a short questionnaire? A before/after example on the landing page would help a lot.

Nice 👍 I hope it scale better

The outlier score cuts research time in half - comparing against each creator's median is way smarter than chasing raw views. Free transcript tool alone removes friction for testing if this fits your workflow.

Smart workflow for turning top Reels into scripts. A few suggestions: support exporting to teleprompter-ready format and adding scene-change markers so editors can cut to the beat. Multi-voice variants (e.g. one script per target audience) would also widen reuse.

I find the idea of identifying outlier videos particularly interesting. Comparing a video's views against a creator's usual performance gives much more context than looking at views alone. I also like how the workflow connects creator research with transcript analysis and AI script remixing. How do you make sure the generated scripts preserve a creator's unique voice while adapting successful content structures to new topics?

The outlier score (views relative to a creator's own median) is a smarter signal than raw view counts for finding what really worked. How does it handle smaller niche accounts with only a handful of videos, where the median can be noisy?

Can the transcript be corrected before AI Script Remix uses it? For Reels with product names or technical terms, a misheard word could change the hook analysis as well as the resulting script. An editable transcript with timestamps would make it easier to check those terms against the source video.