A when to meet survey asks one question, "when are you free?", and its quality comes from four settings you choose before you send it. The dates on offer, the hours of each day, the length of a slot, and the deadline for answers. Get those right and the survey is quick to answer. Get them wrong and you send a second survey.
The four settings
| Setting | Too little | Too much | A sound default |
|---|---|---|---|
| Dates | Two days, and one person is away for both | Four weeks that nobody fills in | The next five to ten working days |
| Daily hours | A two-hour window that suits only you | 7:00 to 22:00, most of it never usable | The hours the group shares |
| Slot length | 60 minutes for a 15-minute check-in | 15 minutes for a half-day workshop | The meeting length, or half of it |
| Deadline | "Whenever" | An hour from now | One or two working days |
Each setting trades the same two things. A wider survey is more likely to contain a time that works. A narrower survey is more likely to be answered. The sections below take the settings one at a time.
Choosing the dates
Start from the date the meeting has to happen by, and work back.
- Leave room for the deadline. If answers are due on Wednesday, do not offer Tuesday. The first date on offer should fall after the deadline, with enough notice for people to plan around it.
- Remove dates you already know are bad. Public holidays, a release day, the day a key person is away. A date nobody can make is a column everyone has to read and skip.
- Stop at two weeks. Most people know their next week well and the week after roughly. Further out, calendars are empty and the answers mean "nothing is booked yet", which is not the same as free.
- Skip days rather than shorten the range. Six scattered dates across two weeks is a smaller survey than ten dates in a row, and it covers the same span.
If the real question is which day, with no hours involved, a time survey is the wrong shape. How to find a date when everyone is available covers that case.
Choosing the slot length
Slot length decides how many cells each person has to mark. For one eight-hour day:
| Slot length | Cells per day | Cells across five days |
|---|---|---|
| 15 minutes | 32 | 160 |
| 30 minutes | 16 | 80 |
| 60 minutes | 8 | 40 |
Use the coarsest slot that still fits the meeting. A one-hour meeting can be found with 60-minute slots. Use 30-minute slots when people's days are broken into half hours, and keep 15-minute slots for short meetings on a short list of dates.
Two details catch people out.
A slot shorter than the meeting means you read the result in runs. A 60-minute meeting on a grid of 30-minute slots needs two neighboring slots that the same people marked. That is fine, and it finds starts on the half hour. Say the meeting length above the survey, so people mark both halves.
A slot longer than the meeting hides usable time. Someone who is free from 10:30 to 11:00 cannot mark a 10:00 to 11:00 cell, so a 25-minute check-in that would have fitted is lost.
Choosing the hours for a group in several time zones
Offer the hours that overlap, not one office's working day. Write each location's working hours in one zone and take the part they share. If London works 9:00 to 17:00 and New York works 9:00 to 17:00, the shared part is 14:00 to 17:00 London time. Offer that window and nobody has to mark a 6:00 slot.
A third location narrows it again. Add Berlin, one hour ahead of London, with the same working day. Berlin's 9:00 to 17:00 is 8:00 to 16:00 London time, so the part all three share is 14:00 to 16:00 London time.
| Office | Working hours, local | The same hours in London time |
|---|---|---|
| London | 9:00 to 17:00 | 9:00 to 17:00 |
| New York | 9:00 to 17:00 | 14:00 to 22:00 |
| Berlin | 9:00 to 17:00 | 8:00 to 16:00 |
| Shared | 14:00 to 16:00 |
When the shared window is this small, widen it on purpose and say so. Offer 13:00 to 17:00 London time and tell the group that the edges are early for New York and late for Berlin. People can then choose to stretch, which is better than a survey that offers them no chance to.
When there is no shared window at all, as with offices twelve hours apart, a survey will not find one. Decide who takes the early or late slot, rotate it between meetings, and use the survey only to choose the day.
Check the offsets close to the date. Regions change their clocks on different dates, so the gap between two cities can differ by an hour for a few weeks a year.
Use a survey tool that shows each person the times in their own zone. Otherwise say the zone in the question.
Setting the deadline
A deadline does two jobs. It tells people when to answer, and it tells them when the question will be closed.
- One or two working days suits most teams. Long enough for someone who is out for a day, short enough that the first answers are still true when you choose.
- Put it at a time of day, not only a day. "By Wednesday at noon" leaves you the afternoon to choose and send the invite.
- Allow longer when people answer by email. People in other organizations read your message on their own schedule.
- Keep the gap short between the deadline and the first date. Every day that passes, someone's free slot gets booked by something else.
Decide in advance what you will do if someone has not answered by the deadline. The usual rule is one reminder, then choose from the answers you have.
What to write above the survey
Say what the meeting is, how long it runs, when you need answers and what you will do with them. How to ask a group when we can meet has templates for chat and email.
Tell people to mark every time they could make, not only the times they like. A survey filled with preferences shows little overlap.
Three more lines are worth adding when they apply.
- Who must attend. If the meeting cannot happen without two named people, say so. It explains the choice you make later.
- Whether it is in person. Travel time changes what "free" means.
- What to do if nothing works. Ask people to say so in the survey or in a reply. Silence and "none of these" look the same unless you ask.
A worked example
Priya needs a 45-minute design review with six people. Four are in London and two are in New York. It is Monday, and the review has to happen before the end of next week.
Dates. She wants answers by Wednesday at noon, so the first date she offers is Thursday. She offers Thursday, Friday and the first four days of next week. The Friday of next week is a release day, so she leaves it out. That makes six dates.
Hours. London and New York share 14:00 to 17:00 London time. She offers exactly that window.
Slot length. The meeting is 45 minutes. With 60-minute slots she would have three cells a day and would miss starts on the half hour. With 30-minute slots she has six cells a day and can look for two in a row. She picks 30 minutes. The whole survey is 36 cells per person.
Deadline. Wednesday at noon, stated in the message, with a note that she will send the invite on Wednesday afternoon.
On Wednesday the best run is Tuesday of next week, 15:00 to 16:00 London time, with all six free. She books 15:00 to 15:45.
Compare that with the survey she might have sent without thinking: ten dates, 9:00 to 17:00, 15-minute slots. That is 320 cells per person, most of them at hours when New York is asleep.
A form or a grid
A general form builder can ask "which of these times work?" with a list of checkboxes. That is fine for five options. Past that, a general form has two problems. A long list of boxes is slow to fill in, and the results come back as a list of replies that you have to add up by hand.
A tool built for availability solves both. It groups the times by day, lets people mark many at once, and counts the overlap. When to meet: find a time that works for everyone shows the counting on a small example.
If you do use a when to meet form from a general form builder, three habits help.
- Write the date, the start time and the time zone in every option, so each line stands alone.
- Add a required name field, and a final option that says "none of these".
- Keep the list to one screen. If it does not fit, you need a grid.
If you have not chosen a tool, what to look for in a when to meet app has a checklist.
Mistakes that lead to a second survey
- Offering dates before the deadline. The early dates pass while people are still answering.
- Copying your own working day. The window suits you and nobody in the other office.
- Using the finest slot "to be safe". It multiplies the work and people answer carelessly.
- Leaving out the meeting length. People mark a spare half hour for a meeting that needs ninety minutes.
- Changing the settings after people have answered. Earlier answers no longer match the question. If the settings were wrong, say so, start a new survey and close the old one.
- Treating no answer as "not free". You cannot tell the two apart. Chase the missing answers before you count.
The same four settings in Slack
Meeting Mouse asks for these settings in one form. Run /meet with a title in a Slack channel and it asks for the dates, up to 14 of them, the daily window, which starts as 9:00 to 17:00, and a slot length of 15, 30 or 60 minutes, which starts at 30. A day can hold at most 24 slots, so a window of 15-minute slots covers up to six hours and a window of 30-minute slots covers up to twelve.
The survey is then one message in the channel. People mark their slots in their own time zone, and the message lists the best times by headcount. The answer form has an "I can't make any of these" button, so a person with no free slot can say so and still count as having answered. There is no deadline setting, so say the deadline in the channel and close the poll from the organizer's menu when it passes. See how Meeting Mouse works.