Create a working copy of the people who are present, put one participant on each line, and shuffle that copy into a full random order. Keep the original attendee list unchanged, because attendance, consent, pronunciation notes, and access arrangements should not be entrusted to a temporary picker result.
Separate the attendance record from the list you will shuffle
The source roster and the speaking list have different jobs. The source roster records who registered or attended; the speaking list is a temporary set of display names that can be reordered without changing anyone's official record.
For example, imagine a 12-person design workshop. Ten people are in the room at the start, one has warned the facilitator that they will arrive late, and one needs to present first before an interpreter leaves. Keep all 12 entries in the attendance record. Make a second list containing the nine people who can be freely ordered, then place the participant with the time constraint first by agreement rather than pretending the exception is random.
Use the shortest label that remains unambiguous in the room. First name plus last initial may work for a small internal group; two people named Sam need another agreed identifier. Do not paste email addresses, employee numbers, accessibility details, or private notes into a list that will be projected on a screen.
Before shuffling, scan the working copy for blank lines, repeated names, titles that belong on separate lines, and people who have stepped away. A clean list is more important than a theatrical animation: a randomizer cannot know that two spellings refer to the same person or that an empty line represents no one.
Generate the full order with Random Picker
Use the Random Picker when you need an entire speaking sequence, not just one name. Paste one label per line and choose Random Order; the tool shuffles every non-empty line and shows an ordered list.
The separate Pick One action supports optional weights written in the form Name@3, but Random Order does not use those weights. It takes the labels and shuffles them once. For a workshop where each available participant should appear exactly once, use plain labels and Random Order instead of trying to express priority with weights.
Read the result aloud before the first talk begins. This gives people a chance to correct pronunciation, flag an overlooked constraint, or say that they are not ready. If the order is acceptable, copy it into the workshop notes or another approved record. The picker does not provide a save or export control, and reloading the page does not recreate the same sequence.
Pseudorandom browser output is suitable for an ordinary workshop sequence, but it is not evidence of audited fairness. Do not use this workflow for a prize drawing, regulated selection, graded assessment, employment decision, or any process whose rules require reproducibility, witnesses, or a certified method.
Handle late arrivals without rerunning everyone else's position
A late arrival should not force the whole group to lose an order they have already prepared for. Set a rule before the session, such as placing late arrivals after the next break or adding them to the end of the remaining queue.
In the workshop scenario, the eleventh participant arrives after three people have spoken. Add that person to the remaining list according to the announced rule; do not reshuffle the three completed speakers or quietly insert the name wherever it is convenient. If several people arrive together, you can shuffle only that small new group and then append it at the agreed point.
The same principle applies when someone temporarily leaves the room. Mark them as unavailable in the facilitator's notes, move to the next person, and offer a clear return point. Random order is a scheduling aid, not a reason to ignore consent, fatigue, interpretation, caregiving, or technology constraints.
If a participant needs to go first, say that the accommodation was applied before the shuffle. Transparency is more useful than claiming every position was random when one position was intentionally fixed. The group can then evaluate the actual rule instead of a fictional one.
Save the accepted order and preserve the original roster
The accepted sequence needs a durable home outside the picker. Put the final order in the facilitator's run sheet, meeting document, or another location that the team already uses, and keep the untouched source roster beside it for attendance work.
For a short session, the Online Notepad can hold a low-risk working copy in the current browser and download a text file. Its local storage is not an account, cloud backup, access-control system, or permanent archive, so do not place confidential participant data there. Move the approved order to the official workshop record before clearing the temporary note.
Label the records so they cannot be confused:
- Source roster: registered or present participants, maintained by the organizer.
- Shuffle input: disposable display names that were eligible when the order was made.
- Accepted order: the announced sequence plus any documented adjustments.
- Completion note: who has already spoken, who deferred, and where a late arrival was added.
If a name seems missing, compare the shuffle input with the original roster rather than editing the roster from memory. A text comparison can highlight differences, but it still cannot decide whether the person was absent, intentionally excluded, duplicated, or recorded under another name.
Verify the sequence before and after the workshop
Verification should confirm both coverage and consent. Before the first speaker, count the eligible people, count the generated positions, and make sure each display name appears once.
Then check the first, middle, and last entries against the copied list. Ask whether anyone has a timing or access constraint that was not captured. Announce the late-arrival rule and where the accepted order will be visible, without displaying private roster fields.
During the session, mark completed talks in the durable run sheet rather than returning to the picker and drawing again. Afterward, retain only the records required by the event's policy. Clear disposable browser lists on shared presentation computers and make sure the official attendance record was never overwritten by the shuffled copy.
This process does not make every workshop outcome equal. Speaking position can affect preparation time and audience attention, and a pure shuffle cannot correct those effects. It does make the mechanics clear: preserve the source, shuffle only eligible display names, document exceptions, and verify the announced result.
Frequently asked questions
Should I use Pick One or Random Order for a speaking sequence?
Use Random Order when every eligible participant should appear once in a complete sequence. Pick One is better for a single low-stakes choice and can use weights, which are not applied to the full-order shuffle.
Can I add a person who arrives after the order is announced?
Yes, but follow a rule stated in advance. Add the person after the next break, at the end of the remaining queue, or in another agreed place without rerunning completed positions.
Does a random order guarantee a fair workshop?
No. It does not account for access needs, consent, language support, role, or the advantage of speaking earlier or later. Review constraints openly and treat the generated order as a draft schedule.
Will the picker save the order if I reload the page?
Do not rely on it. The Random Picker has no save or export control and a new shuffle can produce a different result. Copy the accepted sequence into the workshop's approved run sheet before leaving the page.
Is it safe to paste the full registration spreadsheet into the picker?
No. Use only the minimum display labels needed for the activity. Keep email addresses, employee numbers, accessibility information, attendance notes, and other private fields in the authorized source system.