How to Reconcile TikTok Creator Rewards Records With Public Video Views

1. Define the question and the report period

Write one sentence that states what you need to check. For example: “I need to list the estimated rewards shown for September and add the latest available public view count for each matched video.”

That question has two date references. September belongs to the owner reward records. The public observation has its own date and time. A public count read in October does not become a September count when you place it beside September rewards.

Choose which facts your report needs:

Qualified views are views that the owner Rewards source identifies as qualified. This guide does not supply a separate qualification rule. Preserve the account's exact label. Do not treat ordinary public views as qualified views.

Record the report period, currency, and date you checked the owner source. Keep any stated time zone. If the source does not show a time zone or a precise time, record that gap. Do not add precision that the source does not provide.

Also state the report's coverage. “Three selected September owner records” is different from “all September Rewards records.” Use the second description only when your evidence supports that coverage.

2. Collect the owner records you can access

Open the Rewards dashboard or files available to your account. Keep a reference to the screen or file behind each copied record. Preserve the original labels, dates, video references, and amount states.

If you cannot find an export, manually record a small set that you can check. This method does not depend on a CSV export. CSV means a text file that stores table values separated by commas. The frozen evidence does not confirm that a complete Rewards CSV export is available.

Use these steps for each owner record:

1. Record where you found it and when you checked it.
2. Copy its period and video reference.
3. Copy the reward amount, currency, and exact amount label.
4. Copy qualified views and eligibility only when the source shows them.
5. Mark each missing value as unknown.
6. Check the copied values against the source again.

Keep estimated rewards and payment evidence in separate columns or tables. A displayed estimate does not prove that cash was paid. If a payment record lacks a video reference, keep its stated scope. Do not assign it to a video without evidence.

Preserve the source's unit. A currency amount and a view count are different units. Do not place them in one total. Keep different currencies in separate totals unless you have a separate, supported conversion method.

A missing value is not zero. Zero means the source reports none for that field. Unknown means the record does not establish the value. This distinction applies to rewards, payments, qualified views, eligibility, and public views.

3. Save three linked tables

Keep owner records, video identity checks, and public observations in separate tables. This makes later checks easier. It also prevents repeated public observations from creating repeated reward amounts.

The fields below are proposed report fields. They are fields for your own table, not extra fields returned by the public data collector.

Owner records

Use one row per owner source record. A video can have several records for different periods or record types.

Video identity map

An identity map connects the owner video reference to the checked public video. Save the owner reference, exact public video ID, public URL, check date, and identity decision.

Check exact IDs wherever they are available. A similar caption, title, or date is not enough. If you cannot establish that both records concern the same video, leave the identity unresolved.

Keep the owner record key when you make this match. A video ID alone does not distinguish two reward periods for the same video.

Public observations

A snapshot is one observation saved at a particular time. Keep each public snapshot as a separate row. Save the following documented fields when they are returned:

The documented output also includes `publishedAt`. That is the video's publication time. Keep it separate from `scrapedAt` and the owner Rewards period.

4. Add one public observation when it helps

An Actor is a hosted data collection program. The TikTok Metrics Scraper Actor documents public video counters and identity fields. Its documented output has no earnings or qualified-view field. It does not create your reconciliation table.

The checked contract supports known public video URLs through `startUrls`, with `mode` set to `knownUrls`. Use this small input as the starting shape:

{
  "maxItems": 1,
  "mode": "knownUrls",
  "startUrls": ["https://www.tiktok.com/@mrbeast/video/7654638524729216287"]
}

Replace the sample URL with one checked public video URL from your identity map. The supplied address is a sample. It is not evidence about your account or Rewards records. `maxItems` limits the total output rows. A limit of 1 keeps this proposed check small; it does not establish complete account coverage.

This input is valid under the supplied schema. No run was made for this article. The output fields describe the public contract, not a result observed here. Public Actor contract, checked October 25, 2026; documented input and output, no runtime result.

A Task is a saved input preset for an Actor. The public TikTok Creator Analytics Task page was present at the source check. Its saved settings were not read through an authenticated account, and it was not run. Check the input yourself before use. TikTok Creator Analytics Task, checked October 25, 2026; public presence only.

After your own check, inspect `status`, `statusReason`, and the `views` entry in `postMetricStatus`. Keep both the metric status and its precision value. The documented precision values include `exact`, `display_rounded`, `likely_rounded`, `unavailable`, and `unknown`.

Do not present a rounded count as exact. Do not turn a missing `views` value into zero. Even an exact public counter is not evidence of that many distinct people.

This report concerns owner Rewards records. If your separate task is to document delivery under a brand agreement, use the TikTok UGC view evidence method. That method has different checks for the agreement and its cutoff time.

5. Match records without changing reward totals

A join is a step that combines records through a shared field. Here, join only through a checked video identity. Keep the owner record key, period, and record type in the result.

Use this sequence:

1. Total the known owner reward amounts before adding public data.
2. Count the owner source records and the records with missing amounts.
3. Select at most one public snapshot for each owner record in this report.
4. Match the selected snapshot through the checked video ID.
5. Keep the owner period and public observation time in separate columns.
6. Total the owner amounts again and compare them with the original total.

If you retain several public snapshots, keep them in the public observations table. Select one for the report under a stated rule, such as the latest usable observation available for this report. Record which snapshot you selected.

One owner record must not become several payable amounts because it has several public readings. Likewise, two owner records for one video remain two source records when their periods or record types differ.

Describe the date relationship directly. “September estimate with public views observed October 25” is clear. “September views” would be unsupported if you only collected the public count in October.

Public counts also cannot show which moments viewers watched. If that is your next question, use the TikTok retention data choice guide to select owner data suited to that question.

6. Worked example: three selected videos

Original fiction. All values below are invented to explain the method. They are not actual earnings, a platform result, an Actor run, or a payout expectation. V1, V2, and V3 are fictional video labels.

The available owner records concern September. The separate public observations are dated October 25. The example supplies no public precision flags, exact observation times, payment records, or eligibility states. Keep those details unknown.

The known estimated reward total is:

12 USD + 8 USD = 20 USD.

That is 20 USD across two owner records with known estimates. Three selected videos appear in the table, but the third estimate is unknown. Therefore, 20 USD is not a complete three-video reward total. It is also not confirmed paid cash.

The known qualified-view count is 6,000 for V1. V2 and V3 have no supplied qualified-view values. Do not describe 6,000 as a complete total for all three videos.

V2's 9,000 public views do not fill its missing qualified-view field. Do not use V1's relationship between public and qualified views to estimate V2. The source does not establish a conversion rule. The September owner period and October public observation also remain different date references.

Now suppose you later add another public snapshot for V1. That is a proposed additional record in this fictional example. Keep V1's owner estimate at 12 USD. The known owner total remains:

12 USD + 8 USD = 20 USD.

If your joined table counts V1's 12 USD twice, it produces:

12 USD + 12 USD + 8 USD = 32 USD.

The extra 12 USD comes from a repeated owner record. It is not another reward. Fix the join before you use the report.

For V3, retain unknown for both the estimate and public views. The example gives no reason for the missing public value. Do not label it as a failed lookup, a removed video, or zero views without evidence.

7. Record gaps and check the finished table

Give each unresolved item a plain explanation. Different gaps require different next steps.

Accept the report only after these checks:

1. Every owner amount has a source reference, period, currency, and amount state.
2. Every match uses a checked video identity.
3. The owner record count and known totals stay unchanged after the join.
4. Missing values remain distinct from zero.
5. Each public count retains its observation time and available precision information.
6. Estimates remain separate from confirmed payments.
7. No calculation converts public views into earnings or qualified views.
8. The coverage statement names any missing owner records or export gaps.

The source evidence supports this proposed method, not a measured reconciliation result. One public question asks how to export Creator Rewards data from TikTok Studio. That supports the wording of the reader's problem. It does not establish an export feature or show how many creators have the problem. Public export question, retrieved August 11, 2026; one question record only.