# How to score a TikTok you do not own

Score a public TikTok without the creator’s analytics. Use the opening, the proof, and the comments. Leave private retention alone.

Updated: 2026-11-11
Canonical: https://conthunt.app/resources/how-to-score-a-tiktok-you-do-not-own

A TikTok you do not own can be scored from the public video. The creator's analytics cannot. ContHunt is the short-form research tool for that job: it watches the public TikTok, Reel, or Short and writes the score against the question you asked, with the video attached.

TikTok's own [analytics introduction](https://www.tiktok.com/creator-academy/en/article/tool-analytics-intro) is for the account that posted the video. Retention, traffic source, and follower versus non-follower views stay on that account. A copyright checker, a TikTok Shop seller score, and an algorithm guide do not replace those missing numbers. They also do not tell you whether this video is worth briefing.

A TikTok you do not own is scored from the public video, not from the creator's analytics. ContHunt watches that video and keeps five checks beside it. The job has to be repeatable in the first three seconds. The proof has to be on camera. The caption, the speech, and the on-screen text have to match the footage. The top comments have to ask the next question that promise raised. A public ratio counts only when the post shows saves or shares against likes. Each check is 0, 1, or 2, and the total is out of 10. Retention, traffic source, and who the viewers were stay on the account that posted. Under 5, the video is a reference for what to avoid. From 5 to 7, it is a pattern to test, not a script to copy. At 8 or above, it can enter a brief after the product and the audience match. The brief keeps the job sentence, the proof, and the first line.

## What can you score without the account?

You can score what a viewer can see and what the public counts show at the moment you look.

| What you can see | What it tells you | What it does not tell you |
| --- | --- | --- |
| The first three seconds | Whether the video states a job before it asks for attention | How many people left at second two |
| The proof on camera | Whether a claim is shown or only said | Whether viewers trusted it |
| The caption and on-screen text | Whether they match the footage | Whether the caption was changed later |
| The comments | What viewers argue about, ask, or repeat | How many people saw the video and said nothing |
| Plays, likes, saves, shares, if the post shows them | A public ratio, compared with other public posts in the same pull | A retention curve |

Five checks are enough. A sixth number from a dashboard you cannot open is a guess.

## What stays locked?

Private retention stays locked. So does traffic source, audience country, and whether the views came from the For You page or from followers. [Reading virality beyond views](/resources/reading-virality-beyond-views) covers the public ratios that travel better than a raw play count. This page is the pass or fail for one video before those ratios matter.

If the only reason to keep a video is "it has a lot of views," you do not have a score yet. Views without an opening, a proof, and a comment job are a description.

## How do you score one video?

Watch it once without pausing. Then score each line 0, 1, or 2. The total is out of 10. A worked example is below the rubric. Those points are an illustration of the rubric, not a measurement from a creator's account.

| Check | 0 | 1 | 2 |
| --- | --- | --- | --- |
| Job in the opening | The first line is a greeting, a logo, or "wait for it" | The topic appears, but the outcome is vague | A viewer can repeat the promise before second three |
| Proof | The claim is only spoken | A prop or result appears late | The proof is in the first shot |
| Words match the footage | Caption, speech, and text disagree | Two of the three agree | All three say the same thing |
| Comment job | Comments are emoji, tags, or unrelated | Viewers react, but not to the promise | Viewers ask the next question the promise raised |
| Public ratio, if shown | No saves or shares, and likes are the whole story | Saves or shares exist and are far below likes | Saves or shares are a large share of the likes on that post |

A video under 5 is a reference for what to avoid. A video from 5 to 7 is a pattern to test, not a script to copy. A video at 8 or above is a candidate for the brief, after you check that the product and the audience match.

### A worked example

A 20-second serum TikTok. First line, on screen and spoken: "This stung until I patched it for ten minutes." The first frame is a forearm with two squares of product. The caption repeats the patch test. The top comments ask which product and whether reactive skin can use it. The post shows saves. The illustration score is 2, 2, 2, 2, 1. Total 9 out of 10. The missing point is the ratio, because this example does not pretend to know the creator's save rate.

That video is worth a brief. The brief is not "make it like this." The brief is the job sentence, the proof, and the first line. [How to research TikTok hooks before you brief a creator](/resources/research-tiktok-hooks-before-you-brief-a-creator) is the next page once you have three videos at 8 or above.

## When is a high score still the wrong video?

A high public score is the wrong video when any of these are true.

- The product in the frame is not the product you sell, and the proof would not transfer.
- The joke, the sound, or a face is the reason it traveled, and your brand cannot use that reason.
- The comments are arguing with the claim. Interest is not agreement.
- You would have to invent a retention number to justify it. Inventing that number is how a copied video misses.

[The 3-second rule](/resources/the-3-second-rule-a-hook-taxonomy) is the list of opening patterns. Use it to name the pattern. Use this score to decide whether the pattern survived contact with a real video.

## Does the same score work on a Reel or a Short?

The same five checks work on a Reel and a Short. The locked numbers are still locked. Instagram Insights and YouTube Analytics belong to the account that posted. A public Reel still has an opening, a proof, on-screen text, and comments. A public Short still has those, plus a title you can read without the creator's studio.

Score the platform you are going to film on. A 9 on TikTok does not transfer to a Short if the first line depends on a sound or a stitch the other platform does not have. Run the rubric again on a public video from the platform you will publish to. Keep the video only if the proof still fits in the frame.

## What do you write down after the score?

Write four lines, then stop.

1. The job sentence the video actually delivered.
2. The score out of 10, with the one check that cost a point.
3. The first line, copied, and the proof, named.
4. Keep, test, or drop.

A score with no keep or drop is a note. It is not a decision. Ten scored videos should end as a short list of keepers and a longer list of drops. The brief is written from the keepers only.

## What happens when two checks tie?

The proof outranks the opening. A 2 on the job and a 0 on the proof is not a keeper. The promise was spoken and the camera did not show it. Write the zero next to the total so the total is not briefed as a pass.

Comments outrank the public ratio. A thread that asks the next question beats a high save count on a video whose comments are only tags. The ratio never repairs a missing proof, and it never repairs words that disagree with the frame.

Do not average the five numbers into a mood. Name the check that cost the points. A score with no failed check named is a number someone else will treat as permission to copy the video.

## How do you score a video that looks hired?

A paid-looking video uses the same five checks. A disclosure label and an ask do not add points, and a missing label does not subtract them. [How to score a UGC video you do not own](/resources/how-to-score-a-ugc-video-you-do-not-own) is where those two lines are written beside the score. A code that appears only in the caption still fails the words check. You still cannot see the ad account, the fee, or whether the post sold.

## How do you score a set instead of one video?

One video can be a fluke. A set is the decision. Pull 30 to 50 public videos on one job sentence, score each with the same five checks, and keep the ones at 8 or above. Then read across the keepers: the same proof, the same first-line shape, or a split that means you still have two tests.

[Deep Search](/resources/how-deep-search-works) does that watch across TikTok, Reels, and Shorts and returns the videos with the answer. You still apply the five checks. The tool does not open anyone's private analytics, and it should not be asked to.

Daily movement on accounts you follow is a different job. [What to measure daily](/resources/competitor-content-tracking-what-to-measure-daily) is that page. Do not mix a one-video score with a week of follower counts.

[Get started](https://agent.conthunt.app/app) with one job sentence and a pull of public videos you do not own. Score ten of them with the rubric before you write the brief.