Typically you both say a greeting then a conversation starts.
> Instead of being polite, you are just making the other person wait for you to phrase your question, which is lost productivity.
Please don't count every second of your life in productivity lost to someone else.
The root of the problem might be frustration with not getting something done, and that needs your reflection.
Or just wake up 3 seconds earlier, and if worried about lost sleep, get to sleep 3 seconds earlier or in the words of Arnold Schwarzenegger "sleep faster."
In the case that both people are present and available in the chat at the same time, sure, it's 3 seconds. If not, that extra "hi" can add latency of hours or more. Days if schedules and availability line up badly enough.
In the delay introduced by "hi", you've created uncertainty and ambiguity for your counterpart. They have no idea whether your "hi" is the prelude to something trivial, or something important, or whether it might might be relevant to work they were about to start.
"Asking to ask" is far from the biggest communication issue in the workplace, but it is bad etiquette and a very easy behavior to correct - so why not just make the barest effort to adjust the way you communicate to better fit the medium? Save the "Hi"s and "How are you"s for synchronous communication like calls or meetings. Chat has a different set of pleasantries.
Sometimes it will go days with just that message lingering "hi, how are you?" until I reply to then after a few hours/a day I finally get what she wanted to ask. It's just annoying, I've told her I much rather receive a "hey, how are you? I need some help with X" or whatever else it is the conversation to be. It's just easier and effective async communication (we live in different timezones).
Maybe the soft return is the key here. When I ask a question via slack it’s a single carefully formatted message, while my most irritating correspondents send multiple messages to convey a single question.
Oh, and don’t paste 80k log files in a message. make it easy for me to respond
I called an australian government phone line recently, i was queued, then informed i could press 1 to register for a callback. I pressed 1. Then I was prompted to say my name, I did that. Then I hung up.
Some time later, I was indeed called back - by a robot. The robot greeted me, informed me it was calling back, and requested that I pass the phone to <my name recorded from earlier>. The robot then queued me and prompted me to press 1 when <my name recorded from earlier> was on the line. Since that was me, I pressed 1. After a bit more time being queued, I was transferred to a human operator.
edit: this seems quite bemusing / irritating, but to give the designer of this callback mechanism a little more credit, it was a phone line that businesses would call -- it would not be uncommon for person A to call from a business phone line, with the callback to the company line being answered by some other person B.
It can just as well go like this:
- Bob writes "Hi" and I get a notification
- I enter the chat and wait for Bob
- 30 seconds later the indicator shows he's typing something
- 30 seconds later, still nothing
- 30 seconds later the typing indicator goes away
- I go away. Eventually Bob follows up with the actual content.
That was frustrating for me, not because of some issues I need to work with, but because Bob kept me on hold while he's trying to figure out what to say.
It's actually much more like sending someone an email that just says, "Hi!"
You'd find that weird and off-putting, wouldn't you? Not putting at least some part of the point of the conversation in the initial email completely violates the expectations for how email works.
Chat works fundamentally the same way. Yes, you can treat it as being synchronous—but until you have established an active synchronous conversation (ie, until you and the other person are clearly online at the same time, talking in real time), proper etiquette should be to treat it as asynchronous, like email, and plan one's opening communications accordingly.
That’s why it’s a shitty thing to do.
Are you ok with not having the conversation?
In some types of work, individuals are supposed to do their own time management, and carve out chunks of time for activities that require sustained concentration, and respond appropriately to occasional requests, which are a mixture of low- and high-urgency. For example, someone who needs to read complex academic papers, but occasionally might be interrupted for an urgent production incident. For them, time management is all about avoiding distractions - like synchronous conversations about trivial matters when they're supposed to be working.
Other areas of work have far less need for uninterrupted blocks of time. A manager doesn't just change tasks every hour, they often also respond to e-mails and chat messages during meetings. For them, time management is all about choosing between overlapping meetings, and managing the length of their queue of work by rejecting and delegating tasks. And of course talking to people is the core of their job.
For the former, sending a 'hi' message without the context needed for them to triage it into urgent or non-urgent means they're interrupted twice instead of once - which is pretty inconvenient.
For the latter, though? A dozen interruptions per hour is completely normal, what's the problem?
Async comms should regularly (and, perhaps, by default) be muted to enhance focus. Let them collect and allow the person to manage their own time. "No hello" allows them to have the question ready for them to address when they process their incoming queues.
Synchronous comms should be used to get immediate attention on an urgent issue that needs to be addressed immediately, or be used for tight, rapid iteration in a discussion (e.g. rapid design discussions). Because it necessarily takes attention away from another task, the activation energy should be higher.
It's still needless overhead. Just because interruption 13 is proportionally less painful doesn't make it free.
Me: "Hello!"
Colleague: "Do you have time?"
Me: "Okay"
User is calling
sigh - Picks up
Me: "Hello?"
Colleague: "wait i need to get my headset"
Me: "Okay..."
Colleague: "Do you know about **"
Me: "Hmm, sort of but i need to look into it. i will get back to you"
spend 20 minutes figuring it out
Me writing in chat: "Okay think i know the answer now: **"
Colleague: "You dont need to look into it. Its working now."
46 points by dfboyd on Jan 23, 2021 | parent | context | prev | next [–] | on: Please don't say just hello in chat (2013) I am the original author of the document this document was based on. It was an internal Wiki page at Google written when I was an SRE. After I wrote the original page, someone put up an internal shortlink at "go/nohello". After I left Google, someone took the Wiki page content and [illegally, since it was Google confidential, simply from being on the internal Wiki], and put it up on the net at "nohello.com".
I wish call centres and companies would explain to their agents that it's not like approaching a customer who has walked into a physical shop where the customer already has the context of who they're talking to. Whereas on the phone I have no idea who it is that has called me, it might be an old acquaintance, it might be a call back I'm expecting about something, or might be a cold call that I want to hang up on as soon as possible (which is most of the time).
So legitimate callers would get the friendliest response from me if they simply started the call with "Hello, I'm calling from <place>, regarding <thing>, am I speaking to <your name>?"
The one that really gets me is when a doctor's office calls me about something, but their outgoing number is different from the number I have to call to reach them, and then they start asking me authenticating questions like my DOB and address.
No, buddy. You just called me from a number I do t recognize. I'm not going to give you identifying information!
Then they offer to give me a number where I can call them back, as though that resolves the issue. I'll call you back at your published number, where I always call you, but any old scammer could give me any old number to call them back and it wouldn't make it legitimate.
I just wish they would stop training all their patients to be scam victims.
Also, for me it seems like a terrible strategy as I generally feel like responding to the greeting with: "Well, actually quite annoyed at this moment due to the appallingly poor way you've just started this call."
Sometimes after posting an explanation that has taken you time to think and write, you want to know if the other person understood it and if this is enough, and also if you can go back to your other stuff. Without an acknowledgement, you would likely keep the subject in your head in case the person needs more from you. With a simple "ok", "thanks", or a thumbs up, you can clear your head and go back to other things.
I often do this and forget to upvote correct SO answers as well. I find a solution, realize that I have to research a bunch of other things, and spawn twenty more tabs and the SO answer gets lost in the pile. What's the point of upvoting SO answers if you haven't been able to confirm that the solution is correct?
I guess I sometimes might do this with chat responses as well. But yeah in that situation something like "Thanks I will try it out!" is probably the appropriate response.
Chances are the colleague was just looking at the jira in their browser, so why wouldn't they copy/paste the link...
But yes, when written please, please, please just use the URL. It is at worst the same if people recognize the number, much easier if not. Plus people can quickly just check it to make sure they remembered correctly. Also shout out to ticketing software that puts the title into the URL so you get the confirmation and reminder without even clicking the link.
GitHub and GitLab are awful here. Their default merge and squash messages reference tickets with stupid short forms like #123 or other/project#123 which aren't clickable except in their web-UI. It feels like a form of lock in. I would much rather that they just put a full URL so that I can click it when reading a Git log.
Do that enough times and eventually people start to get the hint.
It didn't changed anything, I guess it's like others mention, it's something that some people have been done since childhood, and it would be very difficult for them to change that.
I now just simply ignore any "hi" or "hello" message as if I've never received them, and only reply actual questions.
To each their own.
The whole point of this is that it is being done in an asynchronous medium, where the participants may never actually be online at the same time, but can still communicate just fine.
- Hi
- Hi
- Do you know what time it is?
- Yes
- Can you tell me?
- Yes
And so on ad infinitum.
And I hate to have to take my mouse, click on the notification, read "Hello", answer "Hello" and wait 30 seconds watching the "is typing" dots...
I'm thinking of writing a macro (I have many) that answers "Hello" to the last message with a key combination. That way, no interruption if I was typing. A new notification will happen when the actual question arrives.
A question other than 'Can I ask you a quick question?' God I hate that. You mean two questions, since that was a question already. And why does it always have to be 'quick'? Does no one ever have a meaty question that will take effort to answer?
I know; I'm a crusty old fart. But boy howdy would I use that AI chatbot if it existed.
There was a point this had become so pervasive I set my Slack status to "I am currently busy working on a prioritized issue. Please open a jira and then speak to management if the case is urgent." then moved slack to another workspace and muted sound. That was my form of time management and it worked well for me personally, most of the time. One can always check with their manager if they are cool with this.
Tools also affect it. Teams is terrible for “structural”, long-lived team chats (oh, the irony!) like are totally natural in, say, Slack, so tends to push more stuff into meetings and ad-hoc group chats. It’s a really bizarre (and endlessly irritating) design choice. You also can’t “digress” within the same channel like you can in Slack without spamming the main chat.
Worse tools = people use them differently (worse) like calling more and doing more private messaging. Add corporate, division, and individual cultures and preferences, and you can get all kinds of weird stuff going on, including the bare “hello”.
But I think a lot of it is also just individual and regional foibles. Some people feel very strongly that a conversation must begin with meaningless pleasantries or it's horrifically rude. Some people feel like it's rude to just ask you a question without first asking if that's OK—even if that's your role.
We could really do with a lot more explicit communication about communication in our culture, corporate and otherwise, setting boundaries and expectations up front.
So the extended etiquette is: if you don’t have time, it’s ok to not reply.
[1] There’s also the concern that your message appears on someone’s shared screen during a video call.
It can be quite difficult to ‘correct’ that mistake once made, as in order to disable notifications, you often need to open up the chat app itself, which makes all of your conversations flash up on the screen.
Yes, of course you can always just stop sharing, disable notifications, and re-enable, but nevertheless, I always try to actually do the opposite, and send ‘hi’ and then the a follow up message for exactly this reason if I’m sending anything that is confidential or contentious. The difference is that I write the full message first (ctrl + a, ctrl + x, ‘hi’, enter, ctrl + v, enter), so that there’s essentially no delay between ‘hi’ and the content of my message.
You never know.
Just had my message barge into an important presentation last week because the presenter got a new computer with the defaults on.
It seems that both of these problems would be solved if the sender just gets to the point?
There's no reason to add delays on both sides so people can hope they finally overlap and can have a synchronous chat. If that's what you want, just ask for a time for it in the initial message.
Yeah, exactly. So how does someone sending you ANY message make a difference? If you are getting distracted by a "hello" you are getting distracted by ANY message, so you may as well turn your notifications OFF and handle messaging asynchronously - its built into the system itself...
What would be the message you eventually receive about my behavior?
And if you bring up my non-response in the future, and I do not verbally respond, what would your impression be of my communication style?
I just learned to not care. I say "hi" to notify that I'm there and let them type. When an answer is required from me I pay attention, think and answer. Until then I ignore it.
I found it so much easier for me to learn not to be (too) distracted than to lose my mind trying to educate humanity to adapt to me.
you: Hi
co-worker: Hello.
you: I'm working on [something] and I'm trying to do [etc...]
co-worker: Oh, that's [answer...]
The mantis beamed Paladin a hail. Hello. Let’s establish a secure session using the AF protocol.
Hello. I can use AF version 7.6, Paladin replied.
Let’s do it. I’m Fang. We’ll call this session 4788923. Here are my identification credentials. Here comes my data. Join us at 2000.
This is a technical problem masquerading as a social problem, which can be addressed by a technical solution.
Instead of expecting colleagues that only support establishing sessions by hello-ing to switch to an alternative protocol, it is possible to support both hello-ers and no-hello-ers by using a similar trick as found in launchd and systemd to reduce startup time by decoupling dependencies between services with a socket. Suppose service A says it depends on service B, and wants to send some data to service B. Instead of blocking service A being started until service B is up, the service manager can allocate a socket, give it to service A and say "here is your connection to service B" and then defer starting service B until service A actually starts trying to communicate using the socket.
Here's how a similar trick can hide the latency of colleagues that only support hello-ing: we offload the responsibility of establishing the chat sessions to a session manager, which is integrated into our chat client. When a colleague requests to begin a session by 'hello!'-ing, this message can be recognized by the session manager as an attempt to establish a session, suppressed and hidden from the user, and the session manager can automatically respond with 'Hi.' to establish the session. colleague is comforted by receiving the 'Hi.' and can trust that a session is established. When colleague begins transmitting data at the application layer, the data can be forwarded to the user as normal and surfaced by the chat program.
https://www.torforgeblog.com/2017/11/15/read-the-first-four-...
There are great tools on these things that help you prioritize and snooze notifications and ensure yourself some quiet time.
Don't ask to ask, just ask - https://news.ycombinator.com/item?id=39762546 - March 2024 (37 comments)
Please don't say just hello in chat - https://news.ycombinator.com/item?id=36623348 - July 2023 (126 comments)
No Hello: A New Wave - https://news.ycombinator.com/item?id=33906174 - Dec 2022 (57 comments)
How to gently enforce “nohello” to a coworker? - https://news.ycombinator.com/item?id=31578259 - June 2022 (1 comment)
Please Don't Say Just Hello in Chat - https://news.ycombinator.com/item?id=31088433 - April 2022 (43 comments)
No hello – please don't just say hello in chat - https://news.ycombinator.com/item?id=30642052 - March 2022 (77 comments)
Don't ask to ask, just ask (2019) - https://news.ycombinator.com/item?id=30639225 - March 2022 (366 comments)
A better way to say Hello - https://news.ycombinator.com/item?id=30094833 - Jan 2022 (1 comment)
No Hello (2013) - https://news.ycombinator.com/item?id=29978860 - Jan 2022 (67 comments)
Don't Ask to Ask - https://news.ycombinator.com/item?id=29603250 - Dec 2021 (2 comments)
No-Hello - https://news.ycombinator.com/item?id=28682658 - Sept 2021 (8 comments)
Please don't say just hello in chat (2013) - https://news.ycombinator.com/item?id=25881800 - Jan 2021 (350 comments)
Don't ask to ask, just ask - https://news.ycombinator.com/item?id=24259156 - Aug 2020 (101 comments)
No Hello (2013) - https://news.ycombinator.com/item?id=24239880 - Aug 2020 (210 comments)
Please Don't Say Just Hello In Chat (2013) - https://news.ycombinator.com/item?id=19648415 - April 2019 (265 comments)
Please Don't Say Just Hello in Chat - https://news.ycombinator.com/item?id=14868294 - July 2017 (46 comments)
like.. yes my dude, ask the question instead of asking if you can ask the question.
Then they will respond with "can I ask you another question" and you reply again with "you just did".
Finally they'll have to ask "can I ask you another question after this one" to which you can reply "yes".
I disagree with it needing cookies though.
Everyone has their different communication preferences.
Personally I have never felt like the no-hello thing is worth caring about - its chat, just let it happen. It is not like you are trapped there unable to do anything else waiting for them to type their question. Just ignore it until it is worth responding to and get on with your day?
Also, "Hi" is a good way of finding out if a person is present, sometimes you might ask to call them instead of typing your question.
It also might be the case there are multiple people who can answer the question, so you want send out "hi" to a few of them to see who will respond right now.
You can even send this to 3 or 4 people and let anyone who doesn't respond first say you caught so-and-so first and are good to go now. Or just delete the message.
This is part of what makes it such poor etiquette. You've created this ambiguous, open-ended distraction for the other person. You expect them to divert focus away from what they're doing so they can poll the chat channel waiting for you to get your thoughts in order.
At least include some context so they have a better idea of what you're about to ask.
They don't need to poll a channel, most chat applications will pop up a notification when a message appears.
What they can poll is the chat app icon, and to not be AFK in the next few minutes.
What would be the purpose of providing more context before actually asking a question?
If you do post just "hi" in a message, your question should come within seconds after.
I use IM as more informal, (AND/)OR conversations with likely back-and-forth, that may require the other person to be present to be of use.
Oh by the way, I have a follow up question....