TikTok Competitor Video Scoring with a CSV Example
Download eight sample videos as CSV, calculate view growth and engagement, and test how score thresholds change a TikTok review shortlist.
Compare two snapshots of the same TikTok video to measure view growth. Then use views, engagement, and comments to decide which videos to study first. This example uses eight fictional videos from SCTCamp’s Cohort 1 Week 4 demo. Four have a snapshot from 24 hours earlier. The CSV uses neutral video IDs and preserves the original sample counts. Download the sample CSV · Original snapshot JSON
How do you calculate view growth and engagement?
Open the CSV in Excel or Google Sheets. Enter =IF(F2="","",B2-F2) in J2 for view growth. Enter =ROUND(SUM(C2:E2)/MAX(B2,1),4) in K2 for engagement. Fill both formulas down and format column K as a percentage. This example counts likes, comments, and shares. Keep view growth blank when you have no earlier count. In your own sheet, record capture times and compare consistent intervals in the same time zone.
video-001 went from 15,800 to 24,000 views, an increase of 8,200. Its latest 1,900 likes, 240 comments, and 160 shares add up to 2,300 interactions. Divide by 24,000 and round to four decimal places to get 9.58% when formatted as a percentage. video-003 has 48,000 views and no earlier count, so its growth cell stays blank.
What gives a video 10 points?
| SCTCamp sample rule | Points for video-001 |
|---|---|
| At least 10,000 views earns 3 points | 3 |
| At least 4% engagement earns 3 points | 3 |
| Positive growth earns 1 point per full 1,000 additional views, with a minimum of 1 and a maximum of 3 | 3 |
| At least 20 comments earns 1 point | 1 |
| An absent prior snapshot earns 2 points | 0 |
video-001 and video-002 both score 10. The tool breaks ties by latest view count, so video-001 ranks first with 24,000 views. The two points for a missing snapshot give first observations a place in the queue. Keep first observation and actual publication date as separate fields. Default configuration · Scoring code
How do the thresholds change the shortlist?
Start with the default configuration and change one setting at a time. The table shows the results of rerunning the public scoring code on the same snapshots.
| Configuration | Passed the filter | Shortlisted |
|---|---|---|
| Default | 8 | 5 |
min_priority_score=8 | 3 | 3 |
only_new=true | 4 | 4 |
max_action_items=3 | 8 | 3 |
The minimum score decides which videos qualify. The shortlist limit controls how many your team will review. The only_new setting keeps videos with no earlier snapshot. Use publication dates to check when they were posted. The scores help prioritize creative review. Estimating sales, profit, or ad lift requires transaction or experiment data you can verify.
How do you reproduce the results?
Install the repository dependencies using the public instructions and run uv run week04 demo. Open weeks/week04-competitor-watch/output/analysis.json to check matched_count, actions, and the priority_score for each video. After the first run creates a local config.json, change one setting under rules and rerun the demo. Check the default ranking in the browser, then use the competitor monitoring guide to plan your own snapshots. Live demo · Open-source instructions