Proving a Giveaway Draw Was Fair: Exports and Timestamps

Every giveaway announcement says the winner was picked fairly. Almost none of them can show it. When a runner-up questions the result, or a client's legal team asks for the paperwork, "we used a random picker" is where most agencies and creators run out of evidence. A claim of fairness and a proof of fairness are different things, and only one of them survives a complaint.

This guide covers the evidence layer specifically: how to export and freeze the entry pool, how to use timestamps to settle eligibility, what a random draw can and cannot prove afterward, and how to package it all into a short client report. If you want the wider system of folders, registers, and retention, that sits in the guide on how agencies run giveaways for multiple clients without losing the audit trail. This one is about the proof itself.

What you actually need to prove

A draw is fair if three things are true, and each one needs its own evidence.

The pool was complete and fixed. Everyone who entered correctly was in it, nobody who entered late or incorrectly was, and it did not change between the deadline and the draw.

The rules were applied as published. If the rules said one entry per person, keyword required, replies not counted, then the filters used in the draw matched that exactly.

The pick was random. The winner came from a method where every eligible entry had an equal chance, and nobody could steer the result.

Regulators think in the same terms. The ASA's guidance expects a promoter to hold evidence that the winner was selected randomly and that valid entries were included, and it has upheld a complaint against a festival that said it used an online comment picker but could not show how the selection was random. Naming a tool is not evidence. What the tool did on the day, captured at the time, is.

Start with the export

The export is the backbone of the proof, because it turns a live comment section, which changes by the minute, into a fixed file you can point to.

A useful export carries enough fields to reconstruct eligibility without going back to the video. TT Picker's comment export, which comes as a CSV you can open in Excel or Google Sheets, includes the username of each commenter, the full comment text, the date and time the comment was posted, any hashtags or mentions in the comment, and whether the comment was marked eligible under your filters. That set is what lets someone else check the pool later: the usernames show who was in, the text shows whether the keyword was there, the date and time show whether it arrived before the deadline, and the eligibility flag shows what the filters decided. The walkthrough on exporting comments with TT Picker covers the mechanics, and the guide to exporting TikTok comments to Excel covers the spreadsheet side.

Three habits make the export credible rather than just present.

Export before you clean anything. If you delete spam or obvious bot comments from the video first and export afterward, your file no longer shows what entrants actually posted, and nobody can tell what you removed. Take the raw export first, then let the filters do the excluding, so the exclusions are visible in the file.

Never edit the raw file. Save it as exported, then do any sorting or annotation in a copy. The original is your evidence, and an original with hand edits in it is not.

Name it so it cannot be confused. Use the giveaway ID, the word RAW, and the export date and time in UTC, so it is obvious which file is which a year from now.

Fix the pool with a file hash

A file hash is a short fingerprint of a file. Change one character in the file and the fingerprint changes completely. Recording the fingerprint of your raw export, along with the time you recorded it, gives you a way to show later that the file you are presenting is the same one that existed on draw day.

On most computers this takes seconds. On Mac and Linux, the command sha256sum followed by the file name prints the SHA-256 hash. On Windows, PowerShell's Get-FileHash does the same. Paste the hash into the giveaway's report along with the export time, and into the register row if you keep one.

Be clear with yourself about what this proves. A hash shows the file has not changed since you recorded it. It does not show the file was complete when you took it. Completeness is a separate check: compare the comment count shown on the video with the number of rows in the export, and note any gap and the reason. Replies, deleted comments, and filtered comments can all explain a difference, and a noted difference with a reason is far more convincing than an unexplained one or a silent one.

Use timestamps to settle eligibility

Most disputes about who was eligible are really disputes about time. Someone commented at the deadline, someone commented a minute after, someone was in a different time zone and thought they had more time. Timestamps settle those arguments, but only if you set the rules for them in advance.

State the deadline with a time zone in your published rules, not just a date. A deadline of midnight on the 15th means different moments for entrants in London, Karachi, and Los Angeles. Writing it in one named zone, and ideally also in UTC, removes the ambiguity.

Say whether the deadline is inclusive or exclusive. "Entries must be posted before 18:00 UTC" and "by 18:00 UTC" are not the same rule for a comment posted at exactly 18:00.

Check which time zone your export uses. Do not assume. Look at a comment whose posting time you know, compare it with the export, and record the offset you found. Mixing an export in one zone with a deadline in another is the most common way a correct draw ends up looking wrong.

Record three times for every giveaway: the deadline, the moment you took the export, and the moment you ran the draw. The export should be taken at or after the deadline, and the draw after the export. Those three times, in order, tell the story of a pool that was fixed before anyone could have influenced it.

Handle late entries by flagging them, not deleting them. In the export, late comments stay visible with their timestamps and are marked ineligible, so a reviewer can see exactly who was excluded and why. The guide on removing duplicate entries in TikTok draws covers the other main exclusion in the same way.

Record the draw itself

The export proves the pool. The draw record proves the pick. TT Picker shows the selection process happening on screen, and the way to turn that into evidence is to record it. A short screen recording that shows the video being loaded, the filters being set, the eligible count, and the result appearing covers the stretch of time a static file cannot. Keep your system clock visible if you can, since it ties the recording to the timestamps above. The guide to screen recording your TikTok winner pick shows how to capture it cleanly.

Before you record, check that the filter settings on screen match the published rules. The recording will be watched by someone looking for a mismatch, and it is better to find it yourself first.

What a random draw can and cannot prove

Understanding the limits of the proof stops you overclaiming, which is where credibility is usually lost.

TT Picker pairs a cryptographically secure random number generator with an unbiased shuffle method such as Fisher-Yates. That is the class of randomness used in security software, and it has the property that nobody, including the tool's operator, can predict or steer the next result. The mechanics are explained in this look at how comment pickers work and whether they are really random, and the closer view of how winners are selected randomly covers the selection step.

The consequence is one many people miss. A draw that is truly unpredictable cannot be re-run to produce the same winner. If a client asks you to repeat the draw to confirm the result, a repeat will give a different answer, and that is a feature of the method, not a fault. The proof therefore is not repeatability. It is the record of the single run: the fixed pool, the filters, the recording, and the output.

If a project needs repeatability, for a high-value prize or a client with a strict audit requirement, there is a stronger method. In a commit-then-reveal draw, the promoter publishes a hash of the entry list and commits to a source of randomness that does not exist yet, such as a public randomness beacon, before the draw. Afterward anyone can recompute the winner from the published values. Some dedicated provably-fair tools are built this way. It is more work, it is not something TT Picker does, and most TikTok giveaways do not need it, but it is worth knowing the option exists when a client asks for proof beyond a recording.

Build the client report

A client does not want to open a CSV. They want one page that says what happened and links to the evidence. Keep it to a single page with the same sections every time, so clients learn where to look.

Giveaway details: client, giveaway ID, platform, post link, dates, and the time zone used.

Rules: the version that applied, with a link or attached copy. A rules generator gives you a consistent starting point across clients.

Pool: raw rows in the export, number excluded and the reasons (late, duplicate, missing keyword, ineligible account), final eligible count, and the SHA-256 hash of the raw file with the time it was recorded.

Method: the tool used, the filters applied, number of winners and backups, and a sentence naming the randomness method.

Result: the winner's username, the draw time, and the draw recording reference.

Verification: what was checked on the winner, when they were contacted, and the outcome. The guide to verifying a giveaway winner is real lists the checks, and the guide to picking a backup winner covers a redraw.

Attachments: raw export, annotated copy, recording, screenshots of the post and the result.

Sign-off: who ran the draw and who reviewed it, with dates.

Write the plain-language summary last, in two or three sentences a non-technical client can forward: how many people entered, how many were eligible, how the winner was chosen, and where the evidence is. Avoid claims you cannot support, such as that the draw "cannot be rigged." State what was done and let the evidence speak.

When someone disputes the result

A dispute is the moment this work pays off. Work through it in order. Retrieve the rules version and the post capture. Open the raw export and confirm its hash matches the one in the report. Check the disputed entry in the file: is it present, what is its timestamp, and was it marked eligible? Check the filters on the recording against the published rules. Then reply with facts, referencing the specific rows and times, rather than defending the process in general terms. In most disputes, the answer is visible in the export within a minute.

Common ways the proof falls apart

Exporting after deleting comments, so the file no longer matches what entrants posted. Editing the raw CSV in place. Mixing time zones between the deadline and the export. A recording that begins after the filters were already set. Calling a draw random without saying what method was used. Taking the export before the deadline and then letting the pool change. Each of these is easy to avoid and difficult to repair after the fact.

A short checklist

Before the draw: deadline stated with a time zone, rules version saved, export taken at or after the deadline, raw file untouched, hash recorded with the time, row count compared with the comment count.

During the draw: filters match the rules, recording running with the clock visible, eligible count and result on screen.

After the draw: output saved, winner verified, report completed on one page, attachments filed. For the wider process around this, including who holds records and for how long, see the guide to the UK rules for TikTok giveaways for what a regulator expects to see.

The bottom line

Fairness you cannot show is just an opinion. A raw export, a recorded hash, timestamps in a stated time zone, a recorded draw, and a one-page report turn it into evidence anyone can check, and none of it takes more than a few extra minutes per giveaway. Do it once as a habit and every draw you run afterward arrives with its proof attached.