Eligibility evidence: why keep a raw record instead of a screenshot

Published: · Last updated:

Short answer

Eligibility evidence is the record that shows you checked a creator before reaching out and what you saw at that moment. A screenshot does not do that job: its time comes from your device, it is easy to edit, and which screen it came from is open to dispute. Evidence that works stores the platform's raw answer with a time set by the system, only grows by appending, and shows it if anything is changed later.

On this page

Short answer

Eligibility evidence is the record that shows you looked at a creator's status before reaching out and what you saw at that moment. To be useful it has to show three things on its own: when the check happened, what the platform answered, and that the record has not changed since.

A screenshot shows none of these. Its time comes from your device, it can be edited in minutes, and which account and which screen it came from is open to dispute. A spreadsheet kept by hand has the same problem. Evidence that works stores the platform's answer as it was given, with a time set by the system, and shows it if anyone touches it later.

What you need on hand at the moment of a dispute is in what to have on hand in a dispute. This guide steps back: how should the evidence itself be kept, and why.

Why is evidence needed?

TikTok's Creator Network Management Policy makes evidence matter in three separate ways.

The first is defence. The policy treats establishing a management relationship with a creator who is in another agency's network or otherwise ineligible, and poaching other agencies' creators by inappropriate means, as prohibited conduct. Consequences include Health Score deductions, a restriction on inviting new creators for at most 30 days, and deductions from payouts. When the objection "that creator was with another agency" arrives after you wrote, being able to show what the platform said at the moment you wrote is the only concrete basis for your good faith.

The second is reporting. The policy expects agencies to report other agencies' violations when they see them, but it also lists reports without reasonably sufficient proof as prohibited conduct. Its contact section says action may be taken against an agency that submits two or more reports without reasonably sufficient proof or evidence. So when you claim another agency wrote to your creator, the evidence is expected from you too.

The third is honesty. The policy lists submitting false, fabricated, altered, inaccurate, incomplete or misleading documents, files, information or data to TikTok as prohibited conduct. A screenshot or spreadsheet row corrected afterwards, even in good faith, can fall under this. That is why evidence that cannot be changed is not only about convincing the other side, but about protecting you.

What can a screenshot show, and what can it not?

A screenshot shows that a screen looked a certain way. But it does not answer most of the questions asked in a dispute.

It does not prove when it was taken. The file's date is the clock of the device that saved it, and it can be changed. Even if a time shows inside the image, that is part of the screen too.

It does not prove it was not edited. Changing a row's status or a username takes minutes and the result may be impossible to tell apart by eye. The other side's first objection will be "you made this afterwards", and you cannot answer it.

It does not prove it was taken before you wrote. The most important question is order: did the check come first, or the message? The screenshot itself does not carry that order.

There is also a practical problem: a screenshot belongs to a person. An image on an employee's phone leaves with them when they leave the agency.

What should a good evidence record contain?

Who: the creator checked, not only by username but by the platform's permanent account ID. Usernames can change; when "which account was this" comes up months later, the permanent ID answers it.

When: the time of the check, as a timestamp set by the system, not typed by hand. A date is not enough; you need to show you checked and then wrote within the same day.

What result: the answer the platform gave, as it was, not interpreted or summarised by anyone. Not an "eligible" note, but the status Backstage returned.

For which agency: the same creator can be eligible for one agency and not another; the region result is the clearest example. The record should also show which agency account the check was made with.

Integrity: being able to tell if a record was changed afterwards or one was removed from the middle.

Shareability: being able to export only the record for the disputed creator or date range, not your whole list.

What does "raw record" mean?

A raw record is the answer the platform gave during the check, stored without being altered. The table you see on screen is a drawing of that answer; the raw record is the source of the drawing.

The difference is this: storing a screenshot stores how a screen was drawn. Storing the raw answer stores what the platform said, and a readable view can be rebuilt from it whenever you want. The view may change, the source stays the same. In a dispute you hand the other side a readable page; the raw record behind it stays with you, untouched.

How is the record kept from being changed?

Two principles. The first is that records only grow by appending: a new check does not overwrite an old record, it is added as a new one. When a creator's status changes the old result is not deleted; the history stays as it was and today's status is the latest record.

The second is a chained seal. A short fingerprint is computed from each record's content. When it is computed, the previous record's fingerprint is included too. That chains the records together: change a single character in an old record and its fingerprint changes, and the fingerprints of every record after it stop matching. Remove a record from the middle and the chain breaks there.

In practice this means the internal consistency of the chain can be checked. The fingerprints can be recomputed from the start on request; if they do not match, you can see where it broke. If they all match, that only shows the current chain is consistent with itself: change a record and recompute every fingerprint after it, and the chain matches again. To show that history was not changed, the current fingerprint has to be compared with a digest pinned outside earlier. So the chain does not independently prove when a record was made or where it came from; for that, the latest fingerprint has to be pinned somewhere dated, outside the control of whoever keeps the record.

What does evidence not prove?

It is worth being honest: even a well kept evidence record does not solve everything.

It proves what the platform said at that moment, not what the creator's real situation was. It cannot know an agreement between the creator and an agency that was not yet reflected in the system.

It does not prove the moment you wrote on its own. The time of the check is in the record; the time of the message is on the messaging side. Putting the two side by side shows the order. That is why who wrote to whom and when should also be kept in one place; see who messaged whom in an agency team.

It does not make an old check fresh. A check from three weeks ago, perfectly intact, is a weak basis for a message you send today. Evidence gains value the closer it is to the moment of contact; we cover this in how long an eligibility check is valid.

It does not clear a contact made despite a negative result. The record is honest; if you wrote after an "In another Creator Network" result, it shows that too.

And it is not a legal guarantee. Evidence is the strongest support you have in a dispute; the outcome is decided by the platform.

How does TT Panel keep the evidence log?

In TT Panel a creator's status is checked before anyone reaches them, and the check is written to the evidence log at that moment. The log stores the check result in raw form with a time set by the system, and the creator is kept by their permanent account ID. Records are only appended, cannot be edited or deleted, each one is chained to the previous one's fingerprint, and their integrity can be shown by recomputing on request. A screenshot does not count as evidence; the readable view is generated from the raw record whenever you need it.

In a dispute you can take an evidence pack for a single creator or a date range and hand it over; the rest of your list stays with you. A fresh negative result rules the creator out for good, and first contact only starts with fresh, positive evidence. TT Panel does not send invites; the invite goes out from Backstage. Compliance with platform rules stays with the agency; the panel shows the eligibility result and the evidence record for every creator.

Eligibility evidence view: result, date and check time on one line

A short checklist if you keep your own record

Is the time of the check set by the system, or typed by hand? Is what you store the platform's answer, or someone's summary? Can old records be corrected or deleted? Can you tell if a record was changed? Is the creator kept by a permanent ID as well as the username? Does the record belong to the agency, or sit on an employee's phone? Can you export the record for a single creator?

This guide is not legal advice. Always read the current rules on TikTok's own pages.

Contact us with your questions or for detailed information about the system

Write to us on WhatsApp or by email; we answer your questions and explain the system in your own language.

Contact us