How to ask for help at work without waiting all day

A help request gets answered fast when one message carries the goal, what you already tried, and the exact thing you need from the reader. In that order.

Most questions that sit in a channel all day are simply unreadable. The reader cannot work out what you want, so there is nothing to type back, and the message slides down the screen while everyone waits for somebody else to decode it. You do not get to make a hard question easy. You do get to make it legible, which turns out to be most of the gap.

Do not ask to ask

Put the whole question in your first message. An opener like "got a sec?" asks the reader to agree to something they cannot see, and the careful ones handle that by saying nothing at all. It is the same self-contained-message rule that makes follow-ups get answered, and in a channel it wastes more to break.

You send "got a sec?" at 4:40. They see it at nine the next morning and say sure. You send the real question at 9:05. Two messages, one day, and no villain in it.

If the question turns out to be too big for chat, the reader will tell you, and they will tell you much faster when they can see its shape.

Say what you already tried

List what you tried, because the first thing any helper does is run through the obvious fixes in their head, and without that list they will spend their whole answer on things you ruled out an hour ago. Then you have to correct them, politely, which takes another round.

A request that gets answered on the first pass usually carries four things:

• The goal, meaning what you want to be true when this is done.
• What you did and what happened instead, in the words the screen actually used.
• What you have already ruled out.
• When you need it, if there is a real deadline.

That second one is where most requests leak. Paraphrasing an error message is the most expensive kind of politeness. "It says something about permissions" gives the reader nothing to search and nothing to recognize, while the literal line is often enough for someone to answer on sight, from a phone, without opening anything.

The same goes for the version of the story you have tidied up. Do not send the clean account of what you did. Send the ugly one, including the part you suspect was silly, because in my own stuck hours the silly part is almost always the part.

Attach your best guess, even when it is wrong

Ask with a wrong answer attached. It feels backwards, and it works better than a clean open question, because correcting is cheap and composing is not. Someone who would never write you three paragraphs from scratch will happily type "no, it is the second one" while walking to a meeting.

A guess also proves you did the reading. There is a real difference between "how does the export work?" and "I think the export runs off the nightly job, which would explain the stale numbers, but I am not sure it is the same job." The first is a request for a lecture. The second is a request for a yes or a no.

Label the guess as a guess so nobody thinks you have already decided, and then let it be wrong in public. That last part is why the guess is the line people cut. It can be wrong where everyone can see it, and that is the bit people flinch at. A wrong guess gets you a correction. Being easy to help is worth more than being right on the first try.

People pad these messages out of politeness, not laziness. A bare list of facts and a blunt guess feel rude when you are the one asking for the favor, so the ask gets wrapped in apology until it sits three lines deep and the reader skims past it. That is the seam I built Unvent for: keep the draft that is already clear and let the tone do the softening, instead of letting courtesy eat the question.

Ask in the open, not in a DM

Default to the channel. A question in a channel gets answered by whoever is free, while the same question in a direct message gets answered by whoever you guessed, whenever they surface. Your guess is usually the busiest person, because expertise and calendar load tend to arrive together.

A direct message is not the polite option. It quietly books one specific person's afternoon so that you do not have to be seen not knowing something in public. Worth naming, because we make that trade while telling ourselves we are being considerate.

The channel version also leaves something behind. The answer sits where the next person with your problem can find it, which is the difference between solving this once and solving it every quarter. Any expert has answered the same question a dozen times in a dozen private windows, and every one of those answers evaporated.

Direct messages do earn their place:

• The content is genuinely private, like someone's pay or performance.
• One named person owns the thing and nobody else could answer it.
• The channel already had its turn and went quiet.

That third case is a follow-up, and it works the way follow-ups work: one named person, not the room, with the original ask restated so nobody has to scroll.

Writing the question often answers it

Writing the question out often answers it, because you cannot state the goal and the dead ends in order without noticing which assumption is wrong. Plenty of my best questions never get sent.

The last one died on me in about four lines. The Unvent landing page had gone completely black after a deploy. Every file looked right where I kept looking, and I was well into being angry at a browser. Writing the question forced me to list what had actually changed, and the answer was sitting in the list. The pages came from the new build and the small file that tells the browser which of my own files to trust came from the old one. Both halves worked. They had simply never met.

This is the part I would keep if I had to throw the rest out. Being stuck is a fog, and the fog is comfortable in a miserable way, because you can keep poking at things and feel busy. Writing the question makes you say what you were trying to do in one sentence, and a surprising number of problems cannot survive that sentence.

So I write the message either way. If it dies halfway, the answer took me a paragraph and nobody else lost a minute. If it survives, what I send is already the good version, dead ends and all.

Close the loop

Say what worked, in the thread, once it works. One line is enough: what the answer turned out to be and what you changed.

Two things happen. The thread stops being a question and becomes an answer, which means the next person with your problem finds it instead of asking you. And the person who helped learns their minute landed, which is the entire reason they will help you again next month.

A well-formed question is mostly an admission: here is what I wanted, and here is where I got stuck. Writing that down is uncomfortable, which is the real reason "got a sec?" stays so popular. The discomfort lasts a minute. The waiting lasts all day.