Free Giveaway Winner Picker

Turn a prepared entry list into a reviewed draw

Use the free giveaway winner picker to select winners and ordered reserves from a list you prepare. Paste one entrant per line or provide a supported CSV, review which records qualify, and draw only after the pool is settled. The outcome is a set of distinct supplied identities with a downloadable record. You do not need a social login, and your entry list is processed in the browser rather than uploaded for selection.

Preparation is part of the job, not something random selection can replace. Collect the submissions you are entitled to use, apply your closing point, and decide which submission types qualify. A giveaway on Instagram, TikTok, YouTube, Facebook, LinkedIn, Threads, X, Snapchat, or Reddit can use a prepared list here. That does not mean the picker connects to those accounts. It does not fetch comments, inspect followers, collect reposts, read messages, or publish an announcement.

Start with a clear counting rule. One chance per eligible identity and one chance per eligible row produce different weighting when somebody has submitted more than once. A visible account label is also not a verified person: two accounts can belong to one person, and a shared display name can describe unrelated accounts. Choose identifiers appropriate to your source and preserve an external reference when you need to find the original submission later.

A weekend reading-box giveaway provides the example below. Its prepared source begins with twelve rows representing nine identities, before duplicate handling, comment filtering, or exclusions. That starting count is not yet the eligible denominator. The useful sequence is source preparation, import, rule selection, review, draw, and record export. Keep the original source separately so a teammate can understand both the accepted outcome and the preparation choices that determined who had a chance.

Bring names or CSV without importing unnecessary data

For a simple list, place one entrant identifier on each line. Do not paste the whole post description, a screenshot transcription containing interface labels, or a paragraph of comma-separated names and assume it represents separate people. If comments affect eligibility or duplicate decisions, a CSV is the clearer source. Use UTF-8 encoding and include a header row so you can explicitly select the entrant and comment columns rather than relying on their position in the sheet.

Only the columns you select are used. A date, platform label, email address, follower count, or evidence URL in another column does not become a hidden eligibility rule. Keep those details in your private preparation sheet when needed and avoid importing unrelated personal information. If eligibility depends on a cutoff or attendance check, finish that work before the draw. The picker does not interpret ignored columns, visit evidence links, or infer identities from email values.

The input limits are ten thousand rows and two megabytes, with an entrant value up to one hundred Unicode code points and a comment up to one thousand. These are limits of this tool, not limits of a social platform’s comment system. Locally bundled Papa Parse 5.5.3 handles CSV structure. Keep commas, quotation marks, and line breaks correctly quoted by exporting a real CSV from your sheet instead of manually splitting comment text around punctuation.

After import, confirm the selected headers, first and last records, and any names containing accents or unusual characters. Read import errors before continuing. An invalid import keeps the existing source intact, so a failed new file is not evidence that the visible pool has changed. File reading can finish asynchronously; if you change the input or reset during processing, the older revision must not replace your newer work. Always check which accepted source is actually displayed.

Make identity matching and duplicate handling explicit

The picker trims surrounding whitespace from identifiers and applies Unicode NFC normalization. Optional switches remove a leading at-sign and fold case for matching. Those choices help reconcile ordinary collection differences, such as a handle copied with an at-sign in one batch but without it in another. They are not account verification. They do not discover renamed accounts, translate names, remove every accent, or establish that similar labels belong to the same participant across several services.

Check the kind of identifier you actually collected. X distinguishes a username from a display name; importing the decorative name rather than the handle can lose an account distinction. For mixed-platform campaigns, the same handle text on two services is not evidence of common ownership. Prepare unambiguous labels where those accounts should remain separate and retain an external lookup to their source. The browser cannot repair a choice that collapsed different accounts before the list arrived.

Exact duplicate rows are identified by the normalized entrant together with the comment after trimming and NFC normalization. Duplicate removal is on by default and can be turned off. Different comments from the same identifier are not automatically exact duplicates. Repeated plain-list lines have the same empty comment, however, so duplicate removal can collapse them. If identical submissions are intended to earn separate chances, preserve them deliberately and check how many eligible rows remain.

Keep duplicate removal separate from the draw mode. Unique mode still gives each eligible identity one chance even if several distinct comments remain. Entries mode lets retained eligible rows supply additional weight, but a selected identity will not win twice. The duplicates and entry rules guide works through these distinctions in detail. Decide the rule from your entry offer before seeing a result, rather than changing it to favor a familiar participant afterward.

Read the review as an explanation of the eligible pool

Inspect missing or invalid records, excluded entrants, keyword failures, and duplicates before accepting the pool. A missing identifier is not made usable by an interesting comment, and a rejected row is not automatically evidence of abuse. Look for ordinary preparation problems such as the wrong selected column or a broken CSV field. Keep a factual reason for each eligibility decision and correct the source when the submission itself is valid but your transcription is not.

The keyword option is a literal, case-insensitive substring search in the selected comment. ‘BOOK’ can match ‘book club,’ while a short term such as ‘art’ can also occur inside ‘cart.’ It is not a whole-word test, regular expression, answer grader, or tag checker. A mention inside a comment remains text; it does not add the mentioned account as another entrant. Review a few matching and nonmatching comments to confirm the chosen term expresses your intended rule.

In the reading-box example, Isha has three source rows, Noah has two, and seven other identities have one each: twelve rows and nine identities. One Isha row and one Noah row are exact copies removed by the duplicate setting, leaving ten rows and nine identities. Orla’s only comment fails the chosen phrase, leaving nine rows and eight identities. Excluding Pavel’s identity removes his only row, leaving eight eligible rows representing seven identities for the draw.

An entrant exclusion removes every row matching that normalized identity, not just one undesirable comment. The final review therefore needs both counts: eligible rows describe the available entry weight, while eligible unique identities describe how many distinct positions can be filled. Neither is necessarily equal to a post’s displayed comment count. On YouTube, for example, held comments are separate from published comments; resolve the source boundary externally instead of treating a public view as automatic proof of completeness.

Choose equal identities or eligible-row weighting

Unique mode gives every eligible normalized identity the same chance at the next position. In the reviewed reading-box pool, seven identities remain, so each has one of seven chances for the first winner. Isha’s two retained rows do not provide an advantage in this mode. This is suitable when your prepared campaign rule is one chance per account label, but it still does not establish one chance per real person or household without separate preparation.

Entries mode draws against eligible rows instead. Isha has two of the eight remaining rows, so her chance at the first position is one quarter. Each of the other six identities has one of eight. That difference is the intended effect of weighting, not a claim that the more active account is more deserving. If you promised capped bonus entries, prepare the allowed rows before import: the picker does not automatically enforce a daily submission cap or calculate attendance bonuses.

After each selection, the chosen identity and all its rows are removed. If Isha is selected first in this example, six rows representing six identities remain. If another identity is selected first, seven rows representing six identities remain, with Isha still holding two rows. Later chances depend on that path. Do not multiply a first-position percentage by the prize count and present it as an exact overall probability of appearing anywhere among several winners and reserves.

Choose one to fifty winners and, optionally, zero to fifty reserves. Their combined count cannot exceed the eligible unique identities, regardless of how many row-weighted chances remain. Eight eligible rows across seven identities cannot fill eight distinct positions. Reduce the request or fix an incomplete source rather than duplicating names to reach a target. This tool deliberately removes repeated identities between positions; it is not a process for awarding multiple prizes to the same supplied identifier.

Review once, then draw with browser-provided randomness

Review creates a frozen eligible pool for the draw. It is a deliberate checkpoint between changing inputs and selecting names, not an automatic draw triggered by typing. Confirm the source, identity options, duplicate setting, keyword, exclusions, weighting mode, and requested positions. Then use the explicit Draw winners action. A second operator can help by comparing those choices with the campaign entry offer and independently explaining the final row and identity counts before anyone is selected.

The selection uses crypto.getRandomValues with unsigned thirty-two-bit values and rejection sampling to choose an unbiased index. The browser supplies the random values. Rejection sampling avoids favoring some indexes when the full integer range cannot be divided evenly by the number of available choices. For ten choices, for example, the thirty-two-bit range has six surplus values; leaving those outside the accepted range allows the remaining possibilities to map evenly onto the ten indexes.

There is no Math.random fallback. If the required secure random source is unavailable, the picker refuses to draw rather than silently changing its selection method. That describes how indexes are chosen, not an independent audit of your campaign. Random selection cannot establish that your source was complete, that the eligibility checks were correct, or that an operator never tried an earlier draw. The saved record documents the accepted run without becoming tamper-proof certification.

A new draw after an existing result requires confirmation. Treat another run as another random outcome, not as a way to improve the first winner’s appearance or popularity. Changing any input invalidates review, results, and exports immediately, so an altered pool must be reviewed again. If you need to correct a real preparation error, preserve the original downloaded record, document the reason, and save the corrected run separately rather than presenting it as the unchanged first selection.

Use ordered reserves without confusing them with winners

Reserve positions are drawn after the winners from the remaining eligible identities. Because each chosen identity has already been removed with all of its rows, winners and reserves are distinct in both weighting modes. Their order matters: reserve one is not interchangeable with reserve three simply because both are backups. Decide your replacement approach before drawing and keep the rank labels when transferring the result into a private working sheet or contact log.

For the reading-box campaign, two winner positions and two reserve positions require at least four eligible identities. If the second winner declines, the organizer can follow the announced process for contacting the next available eligible reserve. That is a follow-up decision using the saved order, not another random selection. The picker does not detect a declined prize, decide when a deadline has passed, promote a reserve automatically, or notify anybody that they have been selected.

Maintain contact attempts and acceptance separately from the original draw. Record which selected position became available, why it became available, which reserve was approached next, and the response deadline with its timezone. Do not edit the original exported winner to conceal a later change. A small chronological log lets another teammate distinguish the initial random outcome from the current recipients, even if several replacements occur before every prize has been accepted and fulfilled.

Use the reserves and redraw records guide when you need a fuller handoff. It distinguishes an exhausted backup list from a faulty original pool and explains how to preserve both runs when a fresh draw is justified. Account checks, availability, and fulfillment stay in your normal campaign workflow. The useful promise here is a distinct ordered selection and a record of it, not automatic management of every event after the draw.

Export the result while keeping the full record private

Choose Copy result for a compact working message, CSV results for selected positions in a sheet, JSON for the full draw record, or PDF for a readable record with guidance. The full record contains a UTC timestamp, random draw identifier, settings, counts, winners, ordered reserves, eligible pool, and excluded row reasons. Keep it with the original prepared source and external decisions. The PDF footer links to this exact tool rather than merely pointing to a generic homepage.

Read the privacy warning before sharing a full record. Eligible comments and exclusion reasons can reveal information about people who did not win, and the public announcement rarely needs that detail. Prepare a limited announcement from the intended result instead of attaching the entire JSON to demonstrate openness. Keep unnecessary personal information out of imported comments in the first place. Unselected CSV columns are ignored, but text intentionally selected as a comment becomes part of the processed source.

Entrant inputs remain in page memory only. The picker does not save entries in localStorage or cookies, send entry telemetry, or upload the entry list. This also means you should download the record before leaving or reloading the page. Keep files you need in your own private campaign storage; the browser session is not an archive. Imports and exports are revision-guarded, so a delayed operation must not substitute an older source or result after you have changed the current input.

Open the downloaded files and compare names, ordering, settings, and counts with the accepted result before handing them over. Name original and corrected draws distinctly. This free picker is separate from SMTasker’s account software and does not create a giveaway integration. For other preparation jobs, browse all free tools; for ongoing account routines, explore SMTasker separately. A completed draw should leave you with both usable selected names and enough private context to explain how the pool was formed.