Hacker News new | past | comments | ask | show | jobs | submit login
What we get wrong about meetings, and how to make them worth attending (timharford.com)
147 points by 80mph 53 days ago | hide | past | web | favorite | 59 comments

For every meeting I create since it takes up people's time, on google calendar I put 2 things in the description:

1. Purpose

2. Agenda

Purpose makes me think - is this meeting necessary? - and also answers why we are having this meeting.

Agenda gives others context on what to expect within the meeting as topics of discussion, as well as clarifies the format of engagement.

This is a great start and far more than most people do. The other thing that most meetings fail at is having someone run the agenda. Someone who keeps the meeting to the agenda, prevents people from getting off topic, makes sure the discussion is pushing towards the stated purpose. The problem is that no one generally wants to be the "bad guy" and they just let the meeting happen. It is also true that sometimes the meeting might be to review a document but people will show up without having read it. This is a waste of everyone's time. But the meeting runner doesn't want to break social norms and call people out. "So you didn't prepare for this meeting Mike, so your time is more important than everyone else's time who did read the document?"

In the end, it is about culture. You have to decide that these things matter. You can have productive 5 minute stand-ups or non-productive 60 minute stand-ups. You have to build the expectation that the point of all meetings is to get on the same page as a team or to make sure you stay there. Most meetings do fail at this.

It's a similar issue in presidential debates. Someone needs to have the authority that something is off topic or off the rails or whatever, but they have to balance that against things that need to be said. And the choice has political implications, big or small politics.

The whole point of structure in these types of dialogues is to reveal the nuggets worth talking about, not cover them up in favor of following the rules.

It is definitely a skill that the meeting runner needs to have. You don't need to keep it completely choked off but you do need the ability to call a spade, a spade. When someone is just repeating someone else to kiss up, then they can move the conversation along.

We have someone who always brings up some story about some company he worked at 20 years ago and blah, blah, blah. But it isn't a "We solved this at company X by doing Y, should we try that here?", it is more like a 15 minute chest pounding of "Back when I ran the largest logistics company providing dental products to all the Fortune 500 and the government, I was in charge of 5K employees. When we faced challenge X,...... blah blah blah". This is where someone needs to step in. Again, most people don't want to be the jerk and so everyone sits there listening to this person talk about himself and wasting everyone's time.


That was why I liked the CNN debates in July. (Democratic presidential primary candidates in the US.) The moderators kept the debate on topic and politely (as best as they could) cut off the candidates when they went off topic.

The next day, a lot of the political coverage complained that the moderators got in the way too much! Sigh.

Agreed, I state this as you need at least:

1. Agenda 2. Expected outcomes

Agenda forces the host/chair to consider how many items are going to be discussed, how long each might take and most importantly scopes the meeting allowing the host/chair to keep things on-topic. If you're organising an important meeting with lots of topics to discuss, but only have 30 mins scheduled due to "scheduling issues", you're doing it wrong.

The expected outcomes ensures that everyone knows up front what is expected to be the output of the meeting. Is a decision is to be made? Are we just looking to gather input?

This helps ensure that the right people are in the room. No-one should be able to say "I'd like to say yes, but Bob is out and I can't say yes without Bob's say-so." It also helps with a focus for the ending of the meeting.

yea agree 100%, usually the last item on my agenda is "make a decision, can agree to disagree"

I started declining small meetings that don't meet this bar. First, I'll reply with, "what's your agenda." If I get a flippant answer, I decline. It pisses my boss off, but it's extremely effective.

Flippant answer: "I'd like to talk to you about [product / project / job title]"

I would like to do something like that, but I don't feel like I could get away with it.

I'd get a follow up email with something like "You have to come. It's your job."

How do people pull off stuff like this?

You could try reversing that and asking them to set an agenda before you a accept the meeting. Keep asking specific questions if needed. If they won't answer, you can pull the "it's your job" card on them.

I make no warranty on this method however, just a thought.

Just ignore the meeting.

How do you have a meeting that doesn't take up people's time?

An hour-long meeting takes an hour, but "takes up people's time" is something people say about meetings if e.g. half the attendees get something useful out of the first half hour and the other half of the attendees get something useful out of the second half hour.

It's a judgment, and if you can engage with the details of what went right and wrong with this week's meetings, then it's possible to tune next week's so the attendees judge them to be "useful" instead of something that "takes up people's time".

edit "that" to "since it" for clarity

Isn't there some fundamental mathematical structure here which makes these problems unsolvable?

Like imagine 20 robots all trying to do some task, each one learning information as it goes. How much time should they spend doing the task vs how much time should they spend communicating with the other robots? (Assuming they are restricted to one or the other)

It's not really a solvable question. It's possible some other robot has a piece of information that would make you much more efficient, however you can't know if they have that information without asking them.

It's possible that taking the piece of information you just learned and gathering all the robots and sharing it with them will help many of them become more efficient, however you don't know if that's true unless you try it.

How can one of the robots work out the value of a piece of information to another robot? Surely all you can do is estimate?

Clearly 0% communication and 100% communication are terrible ideas. However working out which pieces of information to communicate and when is not particularly optimisable, you just have to muddle through as best you can.

> Isn't there some fundamental mathematical structure here which makes these problems unsolvable?

No. Meetings are easy to fix, there's just three simple steps:

a) There must be a point-by-point agenda before the meeting.

b) During the meeting there's somebody to take notes and keep the meeting on topic of the agenda.

c) After a meeting there must be a post-meeting report.

Unfortunately doing it right is a bureaucratic bummer that most people don't like.

Structured meetings only communicate information collectively known to be important. Information you didn't know needed to be communicated is often the most important part of meetings, so having meetings meant to be "derailed" is useful.

> Structured meetings only communicate information collectively known to be important.

Meetings with rules of order aren't meant for communicating information (that's what documents are for), their purpose is to make decisions efficiently.

That's why you have an agenda (what will be discussed and decided), have a chairperson (to ensure the meeting remains on-topic and follows the rules) and keep minutes (so there is a record of what was decided).

When someone comes with new information ("derails") it's perfectly acceptable to take that into account and move a replacement motion, or push it into the next agenda, spin off a delegated group etc. Any widely-used rules of order include numerous mechanisms for dealing with new information arising during the meeting.

But the general point is that meeting is only competent to decide what is in the agenda and decisions only exist if they are in the minutes. If you don't have that ironclad rule than everyone will need to turn up for every meeting, just in case a decision they care about is made. And then turn up for every meeting ever after, because no meetings are final. And argue about what was actually decided.

Rules of order get a bad rap for the same reason as static typing and relational databases: they seem stuffy and slow. But they exist for a good reason and in the long run they are going to be more efficient than making it up as you go.

> their purpose is to make decisions efficiently.

Just underlining this a few times; a meeting is a forum for decisions to be made. If a meeting is used to communicate information then that is a very poor use of meeting time. In some cases, one person is making the decision without input and giving stakeholders an opportunity to stand up and object now; but that is subtly different from communicating information.

Honestly I was almost tempted to copy Jacques' entire comment for emphasis; the rules of order are all critical to making decisions and are all essential to a good meeting. Particularly having an agenda.

That’s really a subset of useful meetings.

Documentation is great to communicate information when it’s clear what needs to be communicated. It falls down when the context of that information is not clear. Handoff meetings are valuable because they reduce the latency between questions being asked and answered.

An email chain over 3 weeks may not take a lot of people’s individual time, but sure eats into the schedule.

In those cases I like to structure de "derails" too.

A problem with derail is when everybody is discussing subject X and it derails to "oranges" because someone just remembered to talk about it in that moment, and is afraid of not talking and forgeting it latter.

If the subject is "X" and it derails to "X1", well, it would be very poor management to not this discussion arise.

Participants are also responsible for keeping the subjects ordered. When preparing for a meeting (yes, everybody in the meeting should have prepared), after receiving the agenda, just see if there is some important topic that isn't covered and warn the organizer. It's very simple to update the agenda previously stating that there is one more subject.

Also, it the meeting organizer doesn't already do it, remember him to leave 10 minutes at the end to "uncovered subjects", where it will be discussed briefly (and will have it's owm meeting if necessary)

I think this applies to communication in large teams in general. There is an amount of shared state that is quite difficult to keep syncronised among different teams. So each team maintains their own version of reality and trys to synchronise it at a particular interval. And a meeting is just one method of syncing.

People realise that the communication overhead is getting out of hand so require longer periods between synchronisation. But this lead to even more discontinuity.

Meetings are not for communication.

They are for making decisions or orientation.

We have plenty of asynchronous ways to communicate with a group.

We also have plenty of ways to set expectations about the agenda and preparation required for a successful meeting.

You´re right. Meetings are bound by network communication constraints. It is a solvable problem in that it´s possible to compute how much time the meeting has to share information, and that helps avoid the problem of having an information collapse due to too many participants, but in the limit, networks and meetings have to learn to self regulate in terms of optimizing use of what is actually a very limited resource.

Judging by human behaviour at most meetings, this is still a work in progress for our species.

Rules of order are a locking protocol controlling who may project sound into the air. They work pretty well.

There is: Amdahl's law. Given a task, the maximum speedup by parallelization is limited, and dictated by its serial component.

It's a simple model, it does not include extra workload due to synchronization or extra workload due to setup cost of nodes (personnel integration, for example). However, it largely explains why teams become less efficient as they grow.

Meetings are just one symptom.

One thing I immediately noticed about meetings at Google is there'd invariably be a shared google doc for the agenda and meeting notes. There's not just one person taking notes trying to capture what happened, everybody just jumps in and updates the doc all at the same time. Huge time saver and a revelation compared to what I'd seen at previous jobs. I've since introduced this idea elsewhere to great effect.

any suggestions on what to use if Google Doc isn't permitted for use within an Enterprise? What tool and configuration can enable this?

If seen whiteboards used for this, by people not knowledgeable enough to setup something like Doc (or rather, not organized enough to set it up in advance).

Then someone snaps a picture and sends it around.

Some of the wiki tools like Confluence now have collaborative editing capabilities. Failing that, you can have someone spin up a copy of Etherpad and have everyone hop on it.

How do you do this effectively so multiple people are not taking notes on the same things? Or is that the point?

Is there a blog post or anything about this approach?

Typically nobody takes notes about what someone else is talking about. Instead the person who has the information enters it directly at the beginning of the meeting. Typically the agenda will be divided into topics, and typically one or two people are most strongly associated with each topic, and that person typically has something to say (and record) for that item. Other people might have questions or suggestions that they can add there as well. So people naturally don't step on each other because they're not trying to take notes on what other people are saying, but instead record what they want to get across to other people in the meeting. It works really really well, as the burden of recording what other people say disappears almost completely, and each bit of information is naturally recorded by the person best suited to record that particular bit accurately. It works especially well for weekly staff meetings where you kind of "go around the room" and get status from each person on what they're working on (often, the verbal status is a very brief summary of the longer version that's recorded by that person in the notes), but it works pretty well for more focused meetings as well.

There's a much more simple answer. You can see where others' cursors are. Don't edit something someone else is about to edit.

This is one of the most succinct ruminations on meetings and I agree with almost every point. Meetings are obviously useful, but they are difficult because there are many ways to do meeting poorly and few know what steps they need to take to ensure meetings are good.

I’ve often thought that attaching a price to meetings would help prompt the question of why this meeting is worth $X of the company’s dime. At the very least it would decrease needless invitees and give some ammunition to the call for higher value meetings.

In Japan, I’ve heard you have to pay out of your cost center to reserve conference rooms and it helps with this, and with not having people needlessly occupy conference rooms. It’s a win win.

Not only in Japan. I know of at least one big, US based multinational that does exactly that.

There meeting rooms are also tied to the attendees, so that only they can open the door. Disruption goes way down when others can't just grab something from an occupied meeting room.

Alternatively you can have only standing meetings. That usually helps reduce wasted time and stay succinct and effective from my experience.

I wish I could say that was my experience. I had a daily scrum meeting for about a year that had at least 15 people, and everyone would take their turn to talk about each post-it note on the board. After standing for about half an hour, it gets really old. Nobody cared what anyone else was saying other than the one manager. When we tried to break it up, the manager said it was more efficient if we were all there at the same time. It might have even been true - for him.

Just think of such meeting as a paid break: check your phone for emails, catch up with Facebook friends, order some things from Amazon, or just let your mind wander. This way no meeting would be "a waste of time" :)

Personally, I've been working as contractor/consultant for the last 8 years, getting paid by the hour worked, so I grew to like those pointless meetings, because they simply are easy money: they count as work, but require almost no effort on my side.

I think you discount the social signaling that goes on at these daily meetings, which I swear is the point of most of them.

If you're not paying attention, bosses will think less of you. So not only do you have to pretend to pay attention they love when you ask lots of gotcha questions to the person trying to give status. Anything to look smart.

>Just think of such meeting as a paid break

If you are paid salary and having to work any overtime, then it is likely taking up productive time and resulting in more unpaid overtime. If you are hourly, then I think that mentality works fine. I wonder if the different incentives every cause a problem in work places with both types of workers.

I once worked for a team that dropped that practice to not force lesser bodily abled to acknowledge that they have a problem.

The 'they have a problem' might need to be expanded upon. Are we talking health issues related to weight, age, pregnancy, or other factors? Some of those are illegal to discriminate upon.

A standing meeting is one that recurs on schedule. A standup meeting is one where participants stand

But often solving the problem is worth more than $X to the company. It's just that the people running the meeting don't know how to actually achieve their goals.

Wow this article/discussion hits close to home - I work at a startup that's trying to tackle the ineffective/unnecessary meeting problem - we've got a Slack app that syncs with your calendar and sends out pulse surveys after certain meetings, then provides a dashboard for surfacing anonymous feedback and insights.

At first I didn’t really grok the problem - I’m used to working at small companies with small teams where bad meetings are pretty obvious and easy to deal with, but after seeing and hearing all the stories from people at the companies that participated in our beta it’s become pretty clear to me the scope of it - it really is an ongoing unsolved issue.

We’re testing different features and approaches and trying to figure out what works/doesn’t work - we’ve got a feature in the works which is like a “meeting exploder” where if enough people in a scheduled meeting don’t think it should happen it can get cancelled or at least reevaluated - we’re trying to make it sort of fun and easy for meeting attendees to give feedback, while still getting enough data that a host or team lead can see whether their meeetings are actually providing value or even need to happen in the first place etc.

It’s early days for us but we've gotten a ton of good feedback already - we’re still working out the kinks but it’s pretty validating to hear about all these teams and people still struggling with “the meeting problem”. We could always use more companies testing early versions too and providing feedback on what kind of features they like/don’t like so if you're on Slack and it’s something you or your team might be interested in give a shout - I'll post a link too, please don't be mad - https://getmarlo.com

The most important thing a meeting can have is an agenda - what will be discussed and by whom. The second thing is a clock. Almost nothing productive ever happens in the second hour of a meeting (unless a long meeting was planned because of the complexity of the issues). If mission creep expands the needs, schedule a new meeting to discuss the added items. The third is a leader - not to control the content, but to ensure that the participants stick to whatever the subject of the meeting. Once they start telling war stories about their prior employer and why the left, someone needs to gently move things along. In the military, they often do stand ups - where the meeting takes place standing up so that no one gets too comfortable or goes on too long.

Finally, I worked in one office that would put out Playdough and Legos for meetings so that people would have things stimulating their hands while they were listening and talking. This was fun.

In Randy Pausch's famous 'last lecture' he also speaks about organizing more useful meetings. His trick that I found works best: organize a meeting half an hour before lunch time and you'll be sure it won't last a minute longer than planned. Nobody bullshits around while hungry.

The amount of missed lunches, late lunches, and lunches eaten in the meeting suggests otherwise.

I think at best we had a break for lunch. At worst it was "just few more minutes and then we'll be free for lunch"

They do. And then you are not just wasting time, you waste time hungry. Besides people are not dumb and will eat before.

Imo, all these "hacks" including standing up designed to cause physical discomfort amount to ineffective management - sort of collective punishment due to inability to reign over few individuals.

This is not true. I have had someone go over their time before lunch and tell us we simply had less time to eat lunch as a result.

In my company's version of OKRs, I count "number of meetings declined" among my most shining and useful quarterly metrics.

A code review is a "meeting" designed to produce innovative outcomes (i most often find a better way when doing a code review than when planning the code or when writing it. At least I notice what we missed)

But the remote nature of code reviews suggest that at least most "innovative" style meetings do not need to be synchronous - or location specific

A very good thing :-)

I'm reading no more articles that use feminine as the gender neutral pronoun it's nonsensical politically motivated language control.

If one finds the traditional masculine gender neutral pronoun form objectionable to there politically sensibilities then use something gender free, pushing the other way is hypocritical stupidity.

Wait, using masculine as a gender neutral pronoun is normal but you get offended if anyone uses feminine as a gender neutral pronoun? That doesn't sound very reasonable.

IMHO the symptoms of a good meeting are: - the organizer speaks less than half the time - some points are discussed and added in the meeting minutes that were omitted in the agenda - some chaos occurs and the agenda is not fully respected (or the order is not respected).

I've read this article about 4 times in the last few hours, and there's so much to say about it -- partly because it resonates pretty deeply with the product I'm currently building (reclaimai.com, apologies for the shameless plug) but also because I think it flies in the face of some logic that has emerged over the past several years. Namely, that meetings are anathema to productive organizations and that individualized "focus time" is the only way to achieve meaningful outcomes (see: http://www.paulgraham.com/makersschedule.html).

The quote that sticks out to me the most:

> Some people will assert that meetings are creativity killers, and “a camel is a horse designed by committee”. But this is absurd. We’ve all been in conversations where one idea sparks another. And while an individual can write a novel or paint a portrait, solo creativity is no way to produce nuclear fusion or a new antibiotic. In a world full of specialists, complex projects require collaboration. Meetings can and do generate ideas that no individual could have conceived alone. They do not do so automatically, however. (emphasis mine)

Managers, in particular, need to internalize this ideology: if you're responsible for leading a broad group of highly-motivated-but-disparate individuals, your fundamental role is collaboration, analysis, collation, and decisionmaking. Not every instance of collaboration needs to take the form of a meeting, but as the author indicates, "there are many situations in which there’s simply no substitute for a meeting". In other words: managers actually need meetings to push the organization's agenda forward. It's not about more or fewer meetings, it's about more high-quality meetings that drive teams to figure hard stuff out and make informed, efficient, and effective decisions.

This isn't to say that focus time is meaningless or not useful to managers, but just that a manager's job is inherently less to individually make and more to synthesize ideas from smart people. That tends to happen in interactive, real-time forums.

One thing Tim doesn't touch on here is how meetings align to priorities, both for the manager and for the company at large. Or: how does your calendar, the declarative record of where you'll likely spend your time, the oft-hedonic treadmill of your week, actually reflect what's important to you at a strategic level? There's a super interesting article that Mike Monteiro wrote in 2013 (https://medium.com/@monteiro/the-chokehold-of-calendars-f70b...) that touches on this exact topic.

His basic thesis is that if all you do is block out time for "working" or "focus time" on your calendar, that time is inherently more interruptible than the meetings. This is anecdotally pretty spot-on for most people: how many times have you been sent an invite for a meeting that overlaps with your precious working time, with the inviter stating "Well, I saw that you just had 'Working' there, so I figured you were free"? That's the interruptibility of focus time in action.

A better way, IMO, is to actually map all the time on your calendar to real priorities. Don't just put "working" time down. State what you're planning to do with that time as part of the actual calendar event, and make it known to those who would interrupt it. This does two important things for your schedule:

1) It makes you think about your time in a much more fundamentally useful and meaningful way. You'll also probably find that a ton of events just don't map to anything strategic for you or the organization.

2) It signals that that time serves a purpose, not just a catchall for reading email or Slack.

If you're interested in this methodology, I wrote a post about it a few weeks ago: https://blog.reclaimai.com/posts/2019-07-11-how-to-fix-your-...

What we get wrong about meetings is scheduling meetings. If you are blocked because you need information or you need to coordinate with someone, walk over and talk to them. Yes, this will interrupt them, but guess what, a meeting interrupts them more, and it delays the unblocking of you. If it isn't critical, don't interrupt them, wait until you see them free or it's lunchtime or something, or send them an email that they can respond to when they are able, or message them on one of the 8 messaging platforms you are probably both logged into at any given time.

The point is that the idea that you need to have a calendar entry and a room and time set aside to talk for a minute is ludicrous and not something that you see anywhere else besides professionals who are charging for their time.

Just to counter obvious defenses of meetings:

- if it's really worth it to take the time then scheduling it is a wasteful delay. If it's not worth interrupting someone briefly then it's not worth interrupting them later via some kind of calendaring system and designated meeting place. That is just interrupting them with a lot of extra steps.

- if your schedule is so booked that you need a calendar to keep all your meetings straight then you had better be a doctor or a dentist or a car mechanic or a salesman who deals with 'strangers' who are paying you, because if you work in an office and are meeting with your workmates so much you need to use a computer to keep it straight you are basically wasting time and desk space and the office would be more productive if you just stopped being there. You aren't doing anything productive and you are actively taking up at least one productive person's time at any given time (this isn't strictly true, you probably spend your time in meetings with people who also spend all their time in meetings, so you are basically just passing gossip and playing politics all day), so you are literally wasting two full salaries worth of company resources and providing nothing.

- it's not avoiding a disruption to schedule a meeting rather than just tapping someone on the shoulder, when you have a meeting coming up you are reminded by your phone and you are discouraged from digging into actual work because you know there's a meeting in 20 minutes. Then you walk to the conference room and the people using the room before you go over schedule by 10 minutes (or more, they probably started late from the previous people), so you stand around and go get coffee and chit chat and wait for the meeting to start instead of just hashing out whatever the subject is standing in the hallway outside the room. Then you go in the room and half the time some random person is invited and you have to dial into the conference system because they stayed home that day, but you don't want them to hear you had a meeting on subject X and didn't invite them, and finally everyone sits there looking at their laptops and phones, respond to emails and message on slack the entire meting, and they don't pay attention because it's a waste of time, then some round number of minutes later you get kicked out of the room. What would often be a 10 minute chat where the person was being completely productive up to the instant you tapped them on the shoulder has turned into an 1.5 hour long ordeal for 4 people.

- if there is one person assigned to figure something out, they can go chat with people 1-1 and gather information and learn about whatever the subject is, discuss options, then make decisions. This is called 'doing work'. If nobody is actually responsible then you get everyone who might know something plus the the dummy that wants to be in every meeting because they don't know what else to do all day and they all posture and try to show how smart they are and try to introduce their pet projects and ideas regardless of how irrelevant they are to the subject of the meeting. Then since nobody is actually in charge you have to find some terrible consensus solution that won't work and is pointless, so then nothing happens.

> a meeting interrupts them more

This depends a lot on the question.

If we're meeting tomorrow to talk about X, I can prepare for that: look at notes and code I've written, check some data, maybe even write some new stuff for you. All of this can be done at a time of my choosing that doesn't interfere with my other responsibilities.

If you just show up at my desk, expecting an immediate answer, several things can happen. Maybe I know it off the top of my head; if so, I tell you quickly, and you go away happy. More likely, however, is that I need a few minutes to think about your question. I have to dig though my own code/notes/whatever, look some things up, and think about them. This is weird to do with someone watching; I think it often takes just as long, if not longer, and the end result isn't as good.

Some of this can be avoided by saying "come back tomorrow morning" or "I'll look into it and email you", but people often expect an answer right away, especially if they've walked a bit to get to your desk.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact