Methodology

Why this index exists

Public TikTok data is abundant but hard to use. The platform is built for watching, not for reference: you cannot sort creators by region, compare engagement across a hashtag, or see how a sound spread over time. Most third-party analytics concentrate on large accounts and ignore the long tail entirely.

This site exists to close that gap. It indexes public TikTok entities — creators, videos, hashtags and sounds — and presents them as structured pages with consistent statistics, so that a creator, researcher, journalist or curious visitor can get a grounded, sourced answer instead of a guess. We publish this page so that every number on the site can be traced to a stated source and formula.

Data sourcing

Our crawlers request publicly accessible TikTok pages — profile pages, video pages, hashtag listings and sound listings — as a logged-out visitor would. We use only public, logged-out data. We do not log in, bypass authentication or access controls, use private APIs, or collect content that is not visible to anyone on the public web.

We do not run any crawler in response to activity on this site. A page view here never triggers a request to TikTok or any other upstream service. Pages are rendered entirely from our own database and cache, and media playback URLs, where present, are resolved on demand from the source platform and held only briefly in a short-lived cache (30 to 60 minutes).

What we store

We store entity identifiers (such as a video ID or handle), public metadata, and statistics we compute ourselves, together with a timestamp of the last successful refresh. We do not store video or audio files, and we do not persist signed media URLs: playback URLs are resolved on demand and expire quickly. When a page says a video was published on a given date or had a given number of views, that figure reflects the source page at the moment of the last successful refresh — it is not a live feed.

Freshness and refresh tiers

Not every entity is refreshed equally often. We allocate refresh effort by how much traffic an entity receives, so that pages people actually use stay current while the long tail is refreshed less frequently:

• Tier A — the most-visited entities, refreshed roughly every 6 to 24 hours.
• Tier B — mid-traffic entities, refreshed every 3 to 7 days.
• Tier C — the long tail, refreshed every 14 to 30 days.
• Tier D — newly discovered entities, fetched once and promoted as they attract traffic.

Every page shows an “as of” date so you can judge staleness for yourself. If you need real-time figures, use the platform directly; this index is a snapshot, not a live monitor.

How each figure is computed

Totals

Follower, following, like and video counts are the values reported by the public source page at the time of the last successful refresh. We do not adjust or smooth them.

Averages

Averages such as average views, likes, comments and shares are computed across the most recent public videos captured for that entity in our index (up to 30 videos, or fewer when the creator has fewer). An average therefore describes the captured window, not necessarily the creator’s entire history. We exclude videos that were unavailable at the time of capture.

Engagement rate

Engagement rate is computed over the same recent-video window:

engagement rate = average likes ÷ average views × 100

We bucket the result into a qualitative label so pages are easier to scan. The buckets are deliberately broad:

Follower bands

Creator size is described with conventional bands, based on follower count at the last refresh:

Posting cadence and recency

Cadence is estimated from the publishing dates of the recent videos in the index (“posts about once a week”, for example), and recency is expressed relative to the last successful refresh. Both are approximations based on captured videos, not a complete publishing history.

Estimated earnings

Earnings figures on this site are estimates, never measurements. They are produced with a deliberately simple model: average recent views are multiplied by an assumed revenue-per-thousand-views rate (RPM). The assumed rate is stated wherever an earnings estimate appears. This index currently assumes a $4 RPM — the upper bound of the published $2–$4 benchmark range — so a creator averaging 100,000 views per video would show roughly $400 per post.

Real earnings depend on audience geography, ad load, brand deals, platform programmes and many other factors that are not publicly visible, so an estimate can be off by a large margin in either direction. Treat every earnings figure as an illustration of scale, not as a statement about any person’s actual income.

What our numbers cannot tell you

• They are not real-time. Every figure has an “as of” date.
• They are not exhaustive. We index public entities we have discovered and captured, which may not include every video or hashtag.
• Averages can be skewed by a small number of viral videos, and by missing videos we did not capture.
• Estimates are models, not observations. We label them clearly and publish the model here.
• We do not publish information about private accounts or anything not visible to logged-out visitors.

Corrections, disputes and removals

If a figure looks wrong, email [email protected] with the page URL and what you believe is incorrect. We review correction requests and respond within a few business days. Corrections are made by refreshing the underlying data or by fixing the computed figure and are reflected in the page’s updated timestamp.

If you are the creator of an indexed profile and want it removed, you do not need a reason and you do not need to dispute a figure first — see the takedown and DMCA page for the removal process.

Changes to this methodology

When we change what we compute or how we compute it, we update this page and its “last updated” date. Material changes that affect published figures are described here rather than made silently.

Contact: [email protected]