paste any public tiktok url and get past the view count: the hook device, the retention curve's weak beat, the comment bait, the sound choice, plus the tiktok you should film next.
analysis runs in 2mv studio · any public tiktok · retention & rewatch signals · direction, not generated video
tap a video to drop its url into the box above
tiktok distributes almost entirely by interest graph, so a cold account and a 2m-follower account enter the same test pool. the analyzer reads a video against tiktok's own mechanics rather than generic short-form rules.
01
retention is measured against the clip you made.
tiktok compares watch time to your own run length, so a 12-second video with 140% watch time (rewatches) outranks a 60-second video people abandon at 20. the breakdown flags the exact beat where the curve would sag.
02
comments are a distribution engine, not vanity.
reply threads keep a video circulating for days. the teardown names the deliberate gap, the missing detail, the disputable claim, the mispronounced name, that makes people type.
03
sound is a ranking surface.
a trending sound puts the video into a second discovery lane, while original audio builds a format people recognise. the analysis reads which lane the video chose and whether the cut rhythm matches the track.
watch time %
top-tier
beat map of where attention drops relative to run length, and whether the clip is longer than the story it carries.
rewatches
top-tier
whether the last frame flows back into the first, and whether any detail rewards a second pass.
comments
high
the engineered gap: unanswered question, contested claim, or a number people want to correct.
shares & sends
high
the forwardable payload, the receipt, the punchline, the place, the price.
interest signals in-frame
high
objects, setting, spoken topic and on-screen words that tell tiktok which interest cluster to test the video against.
sound choice
medium
trending vs original audio, and how the beat grid lines up with cuts and reveals.
on-screen text
supporting
text hierarchy and timing, plus whether the video still reads with the sound off.
0.0-1.0s
01cold open
action already in progress, subject filling the frame. no logo, no 'hey guys'. the analyzer identifies the hook device and whether the frame reads at thumbnail scale.
1.0-3.0s
02stake
the viewer learns what is at risk or on offer: a price, a claim, a countdown, a mess to be fixed. the most common failure point is a video that never states a stake.
3.0-15s
03information density
cut rhythm holds only if every shot adds something. the breakdown flags shots that restate what the previous one already showed.
15s-end
04payoff + comment bait
the reveal lands, and one detail is deliberately left open so the comment section does the rest of the distribution.
last 0.5s
05loop seam
the closing frame either rhymes with the opening, making the rewatch invisible, or hard-stops. the analysis scores which the edit chose.
each axis returns an analysis, a recommendation and production guidance — so the output reads like a brief, not a report.
01
topic
which subjects are actually pulling views in your niche right now.
02
hook
the first-seconds device, phrasing and framing that stops the scroll.
03
pattern
the repeatable structure a format follows from open to payoff.
04
content flow
beat-level pacing: where tension builds, holds and releases.
05
visuals
shot composition, motion, on-screen text and visual density.
06
audio & music
track choice, sound design and how audio drives retention.
07
viewer psychology
the mechanism doing the work, curiosity gap, status, relief, proof.
08
audience profile
who the video is actually landing with, and why it resonates.
posting three times a day is the growth lever.
volume only compounds when the format works. three weak videos train the interest graph toward the wrong audience; one video that holds attention resets the whole account.
you get shadowbanned for using the wrong word.
there is no per-word penalty to undo. videos are tested individually, and a flat video is usually a hook and retention problem, not a punishment.
you must use a trending sound.
a trending sound adds a discovery lane, nothing more. original audio wins when the voice is the format, and it makes the video recognisable across posts.
short videos always win.
length is not ranked; completion is. a 137-second tour holds when each room raises value, and an 8-second clip dies when the story ends at second three.
deleting flops helps the account.
the interest graph reads what you publish next, not what you removed. keeping the flop and diagnosing it is worth more than deleting it.
the creator dashboard only covers accounts you own. read these six the right way, then use the analyzer to explain what produced them.
video views
plays including rewatches from the same viewer.
a volume number only. never read it without watch time next to it.
average watch time
seconds held per play against your run length.
the retention truth. under 50% of run length means the stake beat is missing or late.
watched full video %
share of plays that reached the end.
the number that decides whether tiktok widens the test pool.
traffic source: for you
how much reach came from recommendation vs follows and search.
high for-you share means the content, not the audience, carried the video, that is repeatable.
shares & saves
forwards and keeps, normalised against reach.
shares predict expansion, saves predict reference value. optimise the payoff frame for shares.
new followers per video
follows attributed to a single post.
tells you whether the video sold the account or only the clip.
ai video analyzer
the hub page: how the analyzer works across tiktok, reels and shorts.
open →2mv studio
the full platform — continuous niche monitoring, patterns and playbooks.
open →2mv tools
every free utility built on the research loop, in one place.
open →blog
guides, trends and original research on short-form and organic growth.
open →paste the video url. the clip is decoded shot by shot, then read across eight axes, topic, hook, pattern, content flow, visuals, audio, viewer psychology and audience, and returned as a beat map with hook directions you can film.
tiktok's own creator dashboard is free but limited to your own posts and stops at metrics. 2mv studio is paid because it decodes any public tiktok at frame level and keeps watching your niche continuously, see pricing for current plans.
in almost every teardown it is one of three things: no stake in the first three seconds, a shot that repeats information the previous one already gave, or a payoff that arrives after the audience left. the breakdown names which applies.
yes, any public video. competitor and niche research is where most of the value sits, and it is the part no native dashboard can give you.
no, because posting time barely moves the ceiling on a recommendation feed. it shifts the first hour; retention and rewatches decide how far the video travels.
they are a weak topic hint. what is spoken, shown and written inside the frames carries far more classification weight, three to five relevant tags is plenty.
no. it outputs direction, angles, hooks, structure, pacing, visual and audio guidance. filming and editing stay with you or your editor.
Start getting full visibility into what makes content go viral.