VERDICT: SKIPBERG SCORE 46/100

YouTube Content Performance Predictor

A SaaS tool that predicts how changes to a YouTube video's title, thumbnail, and other elements will affect engagement before you publish.

▶ WATCH THE SOURCE SEGMENT — My honest review of AI Product Designer backed by Y-Combinator (v0 Users Need to See This)

01 THE IDEA

Inspired by a tweet from a 17-year-old who built an algorithm simulating how thousands of users react to a tweet before posting, Greg envisions the same concept applied to YouTube. The product would let YouTubers input their video title, thumbnail, description, and other metadata, then receive AI-powered predictions on expected views, likes, comments, and subscriber growth — essentially de-risking content decisions before publishing.

The core value proposition is AB testing and engagement prediction for YouTube content, giving creators data-driven confidence in their creative decisions. Features would include title optimization, thumbnail comparison, statistical confidence scores, and AI recommendations. Greg demonstrated building a prototype UI using both Poly and v0 (by Vercel), which quickly produced dashboards showing version A vs version B comparisons, cumulative engagement charts, and AI-suggested improvements — validating that the product concept is technically feasible with current AI tooling.

02 THE NUMBERS

EXPECTED ARR

$150K – $1.8M

INITIAL INVESTMENT

$25K + 400h

MONTHLY BURN

$8K + 80h

AUTOMATION

8/10

COMPETITORS

18 · SATURATED →

SKILLS

Machine learning / AI integration, YouTube Data API knowledge, SaaS product development, UX/UI design, Content creator marketing

03 THE VERDICT

The pain is real and well-validated — millions of YouTubers make high-stakes title and thumbnail decisions daily with minimal predictive data. TubeBuddy and VidIQ have proven the willingness to pay, but neither offers true pre-publish AI simulation of audience reaction. The concept is technically buildable today with LLMs plus YouTube Data API, the initial build is low-cost, and the go-to-market path through creator communities is clear. The main risk is defensibility once incumbents copy the feature, so moving fast and building proprietary engagement datasets early is critical.

ALREADY BUILT — BY THE COMMUNITY

I BUILT THIS →

Nobody has claimed this one yet. Shipped it? Tell the story — every submission is hand-reviewed, and approved builds get listed right here with a link to your product.

04 THE FIELD

+15 MORE COMPETITORS + HEAD-TO-HEAD BATTLE PLANSSIGN UP / LOGIN →