# Analyze TikTok Videos That Use a Sound From a Sample

Group supplied TikTok videos by checked public music IDs, keep title-only or unknown cases separate, and compare comparable-age videos inside that declared group. An Actor is a hosted data collection program. A Task is a saved input preset. Analysts with a known video list can use this method to get a sound-group video table with coverage and a bounded comparison.

## Collection setup

Configure the dataset run using build 0.1.36 of the scraper. Set `mode` to `knownUrls` and replace the whole five-video sample list in `startUrls` with one intended known video URL. Set `maxItems` to 1 and `includeProfileMetrics` to `false`.

The run extracts exact native music fields (`musicId`, `musicTitle`, `musicAuthorName`) along with post metrics (`postId`, `postUrl`, `publishedAt`, `views`, `postMetricStatus`, `status`, `statusReason`, `scrapedAt`). All platform IDs are handled as strings.

The primary OUTPUT format is a run summary object containing execution metrics: `providedUrls`, `contentUrls`, `inputUrls`, `resolvedUrls`, `metricRows`, `noPublicMetricRows`, `blockedRows`, `failedRows`, and `statusCounts`. Reconcile the intended source input against the native row results and run summary.

A current one-video positive control returned non-blank `musicId` "7654638538595519263", `musicTitle` "original sound", and `musicAuthorName` "MrBeast". This proves support for these native source fields on resolved posts. It does not prove a group experiment, audible identity, or sound causality, nor does it perform audio fingerprinting, sound page scraping, or global discovery.

## Define the Creative Comparison

1. Define the creative comparison and known video sample. A reader already comparing clips, sounds, and captions needs to know whether the sound group is valid before attributing differences. Choose one platform and public source window. The Actor does not search all uses of a sound from its music page.
2. Start one known video with mode knownUrls, maxItems 1, and includeProfileMetrics false if profile facts are not required. Replace all Task sample URLs. Check resolved post identity, status, and music fields. Public metadata availability can vary; a null music ID is not a unique silent-sound record.
3. Preserve raw music ID, title, author, and source time with the video. Namespace the key tiktok:musicId. Do not join an Instagram audio ID or song title directly to TikTok's ID. Different platform identities need separate audio evidence outside this scope.
4. Build group states: verified_same_platform_record when known IDs match, title_only_candidate when only labels match, and unknown when required metadata is absent. A shared title may be two different uploads or original sound records. Do not group all null IDs together as one sound.
5. Watch or listen to the intended public clips. Record whether the listed sound is actually audible, the used section, voiceover, edits, and uncertainty. This is manual review, not automatic music recognition. Even one platform sound record can be mixed or used at different positions; metadata identity is not identical perceptual content.
6. Choose a comparable age or topic window for the group comparison. Use publishedAt when available and scrapedAt for observation time. Record unknown ages, and do not mix new posts with old posts as if they had the same exposure. Keep counter precision and missing metrics. This guide only needs a simple within-group descriptive view comparison.
7. Report the group's known video count, available metric count, median, range, and sample limits. Never add all group views and call it total sound popularity unless you have a defined exhaustive frame, which this method does not. Keep excluded, title-only, and unknown cases listed.
8. Use the sound group as one creative control while checking other differences: topic, creator, [hook](/en/articles/tiktok-hook-coding/), visuals, publish age, and source availability. A short list or new test follows; neither high views nor retention can be assigned to the sound from public metadata alone.

## Fictional Worked Example and Grouping Table

Here is an original fictional example to show how grouping works. Fictional known videos V1, V2, and V3 have music ID M1, title original sound, same public sound record, and manual review that the audible base cue matches. V4 has music ID M2 with the same title; exclude it from M1. V5 has a null music ID and matching title; keep it as a title-only candidate, not verified M1. V6 has a null music ID and title; it is unknown, not a shared null sound. The verified M1 group has three videos, not five or six.

### Baseline Sound Identity and Eligibility (Original Fiction)

| Video ID | musicId | musicTitle | Sound State | Age State | views | Eligibility / Notes |
| --- | --- | --- | --- | --- | --- | --- |
| V1 | M1 | original sound | verified_same_platform_record | 7 days | 1,000 | Included in M1 comparison |
| V2 | M1 | original sound | verified_same_platform_record | 7 days | 2,000 | Included in M1 comparison |
| V3 | M1 | original sound | verified_same_platform_record | 7 days | 9,000 | Included in M1 comparison |
| V4 | M2 | original sound | different_record | 7 days | 20,000 | Excluded from M1 group |
| V5 | null | original sound | title_only_candidate | UNKNOWN | UNKNOWN | Excluded from M1 group |
| V6 | null | null | unknown | UNKNOWN | UNKNOWN | Excluded from M1 group |

- Baseline analysis: The verified M1 group has 3 eligible videos (V1, V2, V3). Median views: 2,000; range: 1,000–9,000; sample-only sum: 12,000 views. Source IDs and platform records are kept separate from manual audio verification.

### Alternative Scenario: V3 Age Unknown (Original Fiction)

| Video ID | musicId | musicTitle | Sound State | Age State | views | Eligibility / Notes |
| --- | --- | --- | --- | --- | --- | --- |
| V1 | M1 | original sound | verified_same_platform_record | 7 days | 1,000 | Included in age-matched comparison |
| V2 | M1 | original sound | verified_same_platform_record | 7 days | 2,000 | Included in age-matched comparison |
| V3 | M1 | original sound | verified_same_platform_record | UNKNOWN | 9,000 | Retains M1 identity; excluded from age-matched view math |
| V4 | M2 | original sound | different_record | 7 days | 20,000 | Excluded from M1 group |
| V5 | null | original sound | title_only_candidate | UNKNOWN | UNKNOWN | Excluded from M1 group |
| V6 | null | null | unknown | UNKNOWN | UNKNOWN | Excluded from M1 group |

- Alternative analysis: V3 retains M1 identity but is removed from the age-matched view comparison due to unknown age. The eligible set (V1, V2) yields a updated median of 1,500 views. No result proves that the sound works.

V1, V2, and V3 are all observed seven days after publication under the example declared window. They have exact fictional views of 1,000, 2,000, and 9,000. The median for all three is 2,000, and the range is 1,000 to 9,000. The sum of 12,000 is views for the three checked videos, not global M1 plays, sound usage, distinct viewers, or licenses. V4 has 20,000 views, but it is a different record and does not belong in the M1 statistics. If V3's age were unknown, it would remain in the sound identity list but outside the age-matched view comparison. Then the two exact view values of 1,000 and 2,000 have a median of 1,500, and the changed denominator needs a note. No result proves that sound M1 works.

## Next action

Use the public collection tool to check your target video samples by accessing the [TikTok Metrics Scraper Actor](https://apify.com/chronometrica/tiktok-metrics-scraper) alongside the [TikTok Sound and Music Video Performance Task](https://apify.com/chronometrica/tiktok-metrics-scraper/examples/tiktok-sound-and-music-video-performance).

## Sources

- [TikTok Metrics Scraper Actor](https://apify.com/chronometrica/tiktok-metrics-scraper)
- [TikTok Sound and Music Video Performance Task](https://apify.com/chronometrica/tiktok-metrics-scraper/examples/tiktok-sound-and-music-video-performance)
- [TikTok Marketing Discussion on Sound Performance](https://www.reddit.com/r/TikTokMarketing/comments/1qcxv6a/low_retention_rate_on_my_gangster_snowman_tiktok/)