01
Choose what you have
Paste a script or upload a rough cut. You do not need to create it inside RetentionRail first.
AI Simulator · Beta
Find the weak momentsbefore viewers do.
Screen a script or rough cut before you publish. Get the top fix, the evidence behind it, and an honest confidence level—without pretending anyone can guarantee virality.
YouTube videoYouTube ShortInstagram Reel
Illustrative report preview
This is a fixed example. No analysis runs on this public page.
Cut readiness
72/100
Review before publishing. This score measures fixable readiness, not likely views.
Confidence: medium
Highest-priority edit
The title promises speed, but the opening delays the payoff.
Example evidence: the first visual proof appears after the setup. Move the result into the opening, then explain how it happened.
Next action
Revise 0:00–0:07 and re-screen the cut.
After publishing
Confirm the matching post to verify the result.
How it works
One clear path from draft to decision.
The simulator works on its own. Bring content from any writing or editing tool and leave with a prioritized edit—not a wall of AI commentary.
02
Add the real context
Select the platform and format, then add the planned title, duration, and production stage.
03
Screen before publishing
We inspect the opening, pacing, clarity, visual movement, audio, and packaging. Rough cuts may take a few minutes.
04
Fix, re-screen, verify
Start with the highest-impact edit. After publishing, connect the real result so future guidance can improve.
Evidence, not magic
A useful answer should show its work.
Language models explain the findings and suggest edits. They do not get to invent a probability. Every report records what evidence was available and how confident the system should be.
When creator history, platform data, or verified outcomes are missing, confidence goes down. The product does not silently substitute another platform’s data.
What is in the content
The script, title, thumbnail, transcript, audio, motion, cuts, faces, text on screen, and stagnant moments.
Fit with the creator
For connected YouTube channels, we can use the creator’s own history and retention patterns. Instagram confidence reflects the different analytics available there.
Platform and niche context
YouTube long-form, Shorts, and Instagram Reels are kept separate. Missing data lowers confidence instead of being invented.
What happened after publishing
Verified results are linked back to the pre-publish report. This is the feedback loop used to test and calibrate the product.
Built for real work
Less waiting. Less guessing. Less lost work.
Honest progress
Real upload percentage, real job status, and an elapsed timer—never theatrical fake findings.
Resume later
Long video jobs can continue in the background and return after a refresh.
Private by default
Uploads use private signed storage and are never published by the simulator.
Proof after publishing
Match the published post to the original report and measure what actually happened.
Questions
What creators should know.
Does this tell me that a video will go viral?+
No. No honest product can guarantee distribution before viewers react. RetentionRail identifies preventable risks, shows the evidence, and tells you what to fix first. A readiness score is not a view forecast.
Do I have to use the Viral Engine first?+
No. The simulator is a standalone workflow. You can bring a script or video made anywhere, including a cut downloaded from your editor.
Which platforms are supported today?+
YouTube long-form, YouTube Shorts, and Instagram Reels. TikTok is not shown as supported until its platform-specific data and evaluation are ready.
How long does a screening take?+
Script checks are normally faster. Video screening depends on file size and duration and can take several minutes. You can leave it running in the background and resume the result later.
What happens to an uploaded video?+
The upload is private, is used to produce your report, and is never posted to a social platform. You can delete a completed screening and its uploaded video from the simulator.
What if my channel has little or no history?+
You still receive a directional review based on the available content and cohort context. The report clearly lowers confidence and explains what data is missing.
Publish with a reason—not a promise.
Run the draft, fix the biggest risk, and keep the prediction attached to the real result. That is how the simulator becomes more useful with every verified post.
Start a private screeningStandalone workflow · private upload · no virality guarantee