Can I track a competitor's TikTok video, not just my own?

Yes. When you paste a video URL you tag it as 'own' (your KOL collaboration) or 'competitor'. Both are tracked the same way — the same views/likes/comments curve, comment sentiment, and audience approximation. The label drives the comparison view, which lines your video up against the competitor's with a leader marked for views, engagement rate, views-per-hour velocity, and net comment sentiment.

How often does KOLens refresh a tracked video's metrics?

Every 24 hours by default; you can set a tighter cadence per video (down to hourly) when you start tracking. A first snapshot is captured immediately so the curve has a t0 point, then an in-process scheduler re-snapshots due videos on their cadence. Comments refresh on a coarser cadence than the raw metrics because comment scrapes are heavier.

What does the comment sentiment analysis show?

KOLens harvests the actual comment text (not just the count) and classifies each comment positive / neutral / negative, then shows the distribution as a bar plus a net-sentiment score and the top comments by likes. It's a fast read on whether a sponsored post landed with the audience or fell flat — and on a competitor's video, whether their creative is resonating.

Does video tracking show the viewers' demographics?

TikTok doesn't expose per-video viewer demographics, so KOLens approximates the audience using the creator's own audience profile (the people who follow and engage with that KOL) and labels it clearly as an author-level approximation. If you've run an audience snapshot on the creator it uses that real follower sample; otherwise it falls back to enrichment data.

Can Claude track a video for me over MCP?

Yes. The HTTP MCP server exposes three tools: track_video (paste a URL, it starts tracking and bills one scrape unit), get_video_tracking (returns the full curve + deltas + sentiment + audience), and compare_videos (own-vs-competitor scoreboard). Connect KOLens to Claude as a custom connector and you can say 'track this competitor video and tell me how it's doing vs ours' in plain language.

All posts

June 4, 2026·KOLens teamTikTokVideo trackingCompetitor analysis

Track a Single TikTok Video's Performance Over Time

Campaign dashboards track creators. But the unit that actually wins or loses a launch is the single video. Here is how to instrument one published post — yours or a competitor's — and watch it move.

Quick answer

To track a single TikTok video in KOLens: open Video Tracking, paste the video URL, tag it own or competitor, and KOLens captures a snapshot now and re-snapshots on a cadence — a live views/likes/comments curve, comment sentiment, an audience approximation, and a side-by-side comparison against your own video.

Most influencer tooling tracks creators: followers, average views, engagement rate over time. That's the right altitude for discovery and roster management. But once a campaign is live, the unit that actually decides whether a launch worked is a single published video — the sponsored post itself. Did it climb? When did it peak? Is the comment section warm or hostile? How does it stack up against the competitor's post that's eating the same hashtag feed?

KOLens now tracks the video, not just the creator. Here is the full workflow.

Start tracking a video

1
Paste the URL
Open Video Tracking and paste a TikTok video link, e.g. https://www.tiktok.com/@creator/video/762608…. KOLens resolves it to the underlying video id.
2
Tag own or competitor
Mark the video own (a creator you're working with) or competitor (a post you're benchmarking). The tag drives the two-column layout and the comparison view.
3
Watch the curve fill in
A first snapshot is captured immediately so the chart has a t0 point. From there an in-process scheduler re-captures the video on its cadence (24h by default, tighter if you ask) and the views / likes / comments / shares curve fills out.

The metrics that matter

Every tracked video gets a time series, not just a current number. From that series KOLens derives:

Growth curve — views over time, with 7-day and 30-day deltas so you can see acceleration, not just totals.
Views-per-hour velocity — the average growth rate across the tracked span. A post doing 3,000 views/hour on day three is a very different story from one that flatlined after launch.
Engagement rate — likes + comments + shares + saves over views, computed per snapshot.

Comment sentiment, not just comment count

The comment count tells you volume. The comment text tells you whether the post worked. KOLens harvests the actual top-level comments and classifies each one positive, neutral, or negative, then shows the distribution as a bar with a net-sentiment score and the top comments by likes.

On your own collaboration this is your early-warning system: a technically successful post (good views) with a sour comment section is a brand-fit problem you want to catch fast. On a competitor's video it's creative intelligence — you can read, in aggregate, whether their angle is landing before you copy it.

Audience approximation

TikTok doesn't publish per-video viewer demographics. KOLens approximates the audience using the creator's own audience profile — the people who follow and engage with that KOL — and labels it plainly as an author-level approximation. If you've run an audience snapshot on the creator, it uses that real follower sample; otherwise it falls back to enrichment data.

Own vs competitor, side by side

This is where the own / competitor tag pays off. Pick two (or more) tracked videos and KOLens returns a scoreboard: views, engagement rate, views-per-hour velocity, and net comment sentiment for each, with the leader marked per metric. You stop arguing about whether your post is "doing well" in the abstract and start measuring it against the specific competitor video you care about.

Three ways to use it

The UI

The /tracking page is the paste-a-URL surface: two columns (your collaborations vs competitor videos) and a per-video detail page with the curve, content, sentiment, audience, and comparison.

The API

Everything is reachable under /api/videos with your API key: POST /api/videos/track to start, GET /api/videos/tracked for the list, GET /api/videos/{id}/tracking for the full detail payload, and GET /api/videos/compare for the scoreboard.

Claude (MCP)

Connect KOLens to Claude as a custom connector and three tools light up: track_video, get_video_tracking, and compare_videos. You can say "track this competitor video and compare it to ours once a day" and Claude drives the loop.

Where this fits

Video tracking is the per-post complement to campaign-level tracking and the creator watchlist. Watch the creators on the watchlist; watch the videos that matter on the tracker.

READY?

Try it now — 50 free credits on signup.

Track a video free

Frequently asked

Can I track a competitor's TikTok video, not just my own?: Yes. When you paste a video URL you tag it as 'own' (your KOL collaboration) or 'competitor'. Both are tracked the same way — the same views/likes/comments curve, comment sentiment, and audience approximation. The label drives the comparison view, which lines your video up against the competitor's with a leader marked for views, engagement rate, views-per-hour velocity, and net comment sentiment.
How often does KOLens refresh a tracked video's metrics?: Every 24 hours by default; you can set a tighter cadence per video (down to hourly) when you start tracking. A first snapshot is captured immediately so the curve has a t0 point, then an in-process scheduler re-snapshots due videos on their cadence. Comments refresh on a coarser cadence than the raw metrics because comment scrapes are heavier.
What does the comment sentiment analysis show?: KOLens harvests the actual comment text (not just the count) and classifies each comment positive / neutral / negative, then shows the distribution as a bar plus a net-sentiment score and the top comments by likes. It's a fast read on whether a sponsored post landed with the audience or fell flat — and on a competitor's video, whether their creative is resonating.
Does video tracking show the viewers' demographics?: TikTok doesn't expose per-video viewer demographics, so KOLens approximates the audience using the creator's own audience profile (the people who follow and engage with that KOL) and labels it clearly as an author-level approximation. If you've run an audience snapshot on the creator it uses that real follower sample; otherwise it falls back to enrichment data.
Can Claude track a video for me over MCP?: Yes. The HTTP MCP server exposes three tools: track_video (paste a URL, it starts tracking and bills one scrape unit), get_video_tracking (returns the full curve + deltas + sentiment + audience), and compare_videos (own-vs-competitor scoreboard). Connect KOLens to Claude as a custom connector and you can say 'track this competitor video and tell me how it's doing vs ours' in plain language.