Two public TikToks can be compared without either creator’s analytics. ContHunt is the short-form research tool for that comparison: it watches both videos and keeps them next to the question you asked. A head-to-head of follower counts does not do this. Those tools rank accounts. This page picks a video.
You still cannot see retention, traffic source, or who the viewers were. How to score a TikTok you do not own is the rubric for one video. Use it on both, then use the steps below to decide which one enters the brief.
Two public TikToks are compared by watching both videos, not by ranking the accounts. ContHunt keeps the pair next to the question you asked. Line up the same five things: the job in the first three seconds, where the proof appears, whether the words match the footage, what the top comments ask, and a public ratio only when that post shows it. Three differences pick the winner. The proof that a viewer can see sooner beats a proof after a greeting. Comments that name something a person could buy beat tags and punchlines. Words that match beat a caption the frame never shows. Follower count, posting time, and a higher play count do not pick. If those three differences tie, keep both videos as tests. Do not blend one’s opening with the other’s caption. Account tools are for a hire. This sheet is for the video that enters the brief.
What are you comparing?#
Compare the videos, not the accounts. An account with more followers can post the weaker of the two videos. A small account can post the one whose proof shows up in the first frame.
Line up the same five things on both:
| Line | Video A | Video B |
|---|---|---|
| Job in the first three seconds | What a viewer can repeat | What a viewer can repeat |
| Proof | Where it appears, in seconds | Where it appears, in seconds |
| Words | Speech, on-screen text, and caption agree or they do not | Same check |
| Comment job | The question viewers actually ask | The question viewers actually ask |
| Public ratio, if the post shows it | Saves or shares against likes | Saves or shares against likes |
Leave a cell blank when the post does not show it. Do not borrow the missing number from the other video, and do not invent a retention curve.
Which differences count?#
Three differences are enough to pick. More than three and the sheet becomes a second viewing.
- Where the proof sits. A proof in the first shot beats a proof after a greeting, even when the later video has more plays. The viewer you are briefing will not wait for a payoff the reference itself buried.
- What the comments ask. “What product is that?” is a different result from “this is so funny.” The first comment job can move into a brief. The second usually cannot, unless the joke is the product.
- Whether the words match. If one video’s caption promises a result the footage never shows, it loses to the video whose caption, speech, and frame say the same thing.
TikTok’s creative best practices tell advertisers to put the point in the opening seconds. That is why a proof after a greeting loses to a proof in the first shot, even when the later video has more plays.
Follower count, posting time, and hashtag count are not on this list. Reading virality beyond views is where a public ratio belongs, after both videos have passed the three differences above.
What does a worked pair look like?#
Both videos are about a serum. The numbers below are an illustration of the sheet, not a pull from either creator’s analytics.
Video A opens on a forearm with two squares of product and says the patch test out loud. The caption repeats it. The top comments ask which product and whether reactive skin can use it. Proof is in the first frame.
Video B opens on a morning routine, names the serum at second eight, and shows the bottle at second twelve. The comments are mostly fire emojis and people tagging friends. The plays on the post are higher than Video A in this illustration.
Video A wins. The higher play count on B is not a reason to brief the routine. The brief needs the proof a viewer can see immediately, and the comments that ask the next question.
If your product cannot show that patch test, A is still not your script. It is the standard. Film a proof you actually have, in the first shot, and write the first line the way A did.
What if the three differences tie?#
If the proof sits in the same place, the comments ask the same question, and the words match on both videos, you do not have a winner. Keep both as tests. Do not average them into one script. One video’s opening plus the other video’s caption is a third video nobody posted.
The public ratio breaks that tie only when both posts show saves or shares against likes, and only after the three differences have already tied. A higher play count does not break it. If one post hides the ratio, that cell stays blank. You do not copy the other video’s number across.
What must you leave out of the blend?#
Three blends fail even when both videos scored well on their own.
- The earlier proof with the later caption. The caption was written for a different frame.
- The stronger comments with the weaker opening. Viewers asked about a video you are no longer filming.
- A public ratio from the video you dropped. The ratio belongs to the post that earned it.
Write the dropped video’s first line in the note, and mark it as a line the creator must not shoot. Deleting that line is how it comes back in the next draft.
What does a different product change?#
The pair below is an illustration. It is not a serum, and it is not a pull from a creator’s analytics. The product is a stain stick for a shirt. The job sentence, written first, is that someone with a stain on a shirt should believe the stick can lift it, because they see the stain leave.
Video A opens on the stain leaving the cloth. The speech names the stick in the first sentence. The caption repeats the stain. The top comments ask which stick and whether it works on ink. The proof is the first frame.
Video B opens on a closet, names the stick at the ninth second, and shows the tube at the twelfth. The comments tag friends. The plays on B are higher in this illustration.
Video A wins. B’s play count is not the brief. If your product cannot show a stain leaving, A is the standard, not your script. Film a proof you have, in the first shot, and write the first line to match that proof.
The sheet does not change because the product changed. The same three differences pick. A serum pair and a stain pair can both be true on the same afternoon. They are two briefs. Do not merge the stain leaving into the serum’s patch test because both are “proof in the first frame.”
When is the pair a UGC video?#
If either video shows a paid label, a spoken ad line, or a code, those facts are not a new column on this sheet. How to score a UGC video you do not own writes the disclosure and the ask beside the score. A labeled ad with the proof in the first frame still beats an unlabeled video whose proof arrives after a greeting. You are picking a video to learn from, not a creator to hire.
When is the account tool the right tool?#
An account comparison is the right tool when the question is which creator to hire or which account grew. It is the wrong tool when the question is which video to learn from. Follower location, audience overlap, and a live counter do not show the opening.
Use the account tool for the hire. Use this sheet for the video. Do not paste an audience chart into the brief and call it creative research.
How do you stop at two?#
Start with a pile, not with a pair. Score the pile with the one-video rubric. Keep the two highest that share your job sentence. Only those two go on the sheet. A third column means you have not decided what the job is.
Deep Search can watch the pile across TikTok, Reels, and Shorts. The pair you compare should be on the platform you will publish to. A TikTok that wins on a stitch does not automatically beat a Reel that wins on a caption, because the other platform may not have the stitch.
Write the outcome in one line: “Brief A’s opening and A’s proof. Drop B.” What to measure daily is the page for watching those accounts over a week. It is not this decision.
Get started by pulling public videos for one job sentence, scoring them, and putting only the top two on the sheet.
