Video Score: Test Your Video Idea Before You Create

It predicts how a video idea is likely to perform before you create it, rating your concept 0 to 100 against 10 viral frameworks tailored to your niche, platform, and country. You get the prediction at the idea stage, with no filming or editing required. The frameworks cover the mechanics that decide whether content gets seen: hook strength, emotional triggers, structure, pacing, and trend alignment, each weighted for the context you select. A cooking Short in Brazil is not scored like a business explainer in Germany, because the reference patterns follow your market. The report reads like a to-do list rather than a verdict: it names the frameworks that are strong, the ones that need work, and what to change before production starts. Every score is a probability signal based on current performance patterns, not a guarantee of views.

No. You only need the idea. Describe the concept and PreViral scores it at the pre-creation stage. If you already have a finished video, use Analysis instead. The practical difference matters for your workflow: Score exists so you can filter ideas while changing them is still free, long before any footage exists. Write down the concepts you are considering, score each one, and film the strongest instead of guessing. When a promising idea scores low, the framework breakdown shows whether the hook, the premise, or the platform fit is the problem, so you can revise and score it again in minutes. Analysis then covers the other end of the process: it reviews the finished cut before you upload and returns timestamped feedback. Idea first, cut second, and both checks happen while you can still change the outcome.

The score reflects how closely your idea matches the patterns driving reach in your niche and platform right now. It is a probability signal, not a guarantee, so it tells you which ideas are worth your time before you invest hours. We are deliberately transparent about what that means in practice: predictions are logged in a public track record before the outcomes are known, and we then measure real 7 and 30 day performance, with misses kept visible instead of quietly removed. The engine is built on more than 2,500 hours of research across 57 niches, 5 platforms, and 30 countries, and the reference patterns are refreshed so the score never grades your idea against a stale meta. Treat the number as a strong signal and a concrete fix list, and make the final call yourself.

YouTube Shorts, YouTube long-form, TikTok, Instagram Reels, and Facebook Reels, across 57 niches and 30 countries. Every prediction is tailored to the niche, platform, and market you select. That context matters more than most creators expect, because the same concept rarely performs the same way twice: pacing that holds attention on TikTok can feel rushed on long-form YouTube, and a hook that works in the US fitness niche can fall flat for finance viewers in Germany. The frameworks stay identical across contexts, but their weighting and reference patterns follow the combination you choose. If you publish one concept on several platforms, score it once per platform and compare the results before you commit to an edit. And when your exact niche is missing, the closest match still gives a useful directional signal you can act on.

0 to 30 needs rework, 31 to 60 is average, 61 to 80 is good, 81 to 90 is excellent, and above 90 is viral ready. Each score breaks down which of the 10 frameworks are strong and which need work. Read the weakest dimensions first, because they are usually the cheapest wins: a slow hook or a crowded premise can often be fixed in minutes, and rescoring the revised idea shows you immediately whether the fix moved the number. The bands are designed as a filter for your production time, not as a promise of views, so treat anything above 60 as worth serious consideration and anything below 30 as a concept that needs a new angle rather than better execution. Over time the loop compounds: score, revise, produce, then compare outcomes against the public track record.