We adopted Slack and I hate it but use daily. What I find so frustrating with Slack is that it's
1. Super terrible at threading. Threads are clearly discouraged and don't nest. This appears to be by design, but even Usenet handled threading better.
2. Slack is an information wastebasket - at least in my experience it's really hard to find information once there's a significant traffic in a deployment.
3. Too Much Notification! I have to mute channels that has a high rate of addressing the channel.
However there's no denying that it's popular. Seems like every place I go is using Slack. At least slack allows me to edit my typos.
Slack is a tool, and like all tools it can be used badly. You work at a company that uses it badly.
- No one should be using Slack for deep, branching conversations that require nested threads. Those would be lost and hard to refer back to in any system. Learn to separate and focus on things better.
- Slack is not a documentation system or a knowledge base, and shouldn't be treated like one. Pull salient points out of Slack chats and into a better place for holding information you'll go back to. I genuinely believe that Slack's crappy search tool is a feature. It discourages using Slack history as documentation.
- No one should be using channel wide messaging to the point where people are annoyed enough to mute notifications. Those people should be regulating what they broadcast, and use email instead when things need to be seen by lots of people.
Slack is a great tool for short form conversations, checkins, and automated notifications. But that's all. Using it for something that it's not fit for won't work, but that's not Slack's fault.
Your argument is just another “you’re holding it wrong”
People are using these chat systems for such things so maybe Slack as a company should build in the smarts extract the data into a more permanent format.
I get the _strong_ impression that Slack considers information that leaves Slack as a threat to the business. Thus, I don't think it's a lack of smarts, it's a lack of will
Contrast that to https://zulip.com/help/public-access-option which is one of the major "what the hell" for me when any big open source community picks Slack for their "chat" solution: unless the community has deep pockets, messages both age out and are not indexed by search engines
Company mission statements are marketing, not actual missions. If you want to see the mission of the company look at what it actually does rather than what it says it aims to do.
Did you notice we are saying the same thing?: Slack’s mission was to be a knowledge base but, to date, they have failed to execute.
It’s not because mission statements are marketing bullshit that distract outsiders from an alternate nefarious goal. It’s simpler than that: your mission is where you want to go, but in many cases a business just can’t complete that mission.
I've noticed that a lot of tools, ecosystems and companies that have died out in tech have done so because they tried to adapt the users to their way of thinking rather than vice versa.
Frankly, I think if slack doesnt adapt to some of these use cases it'll die off and be replaced by something that does.
That said, I hate usenet style multi threading in any context. Im not sure why some people like it so much.
It's baffling the degree to which Teams doesn't work on a purely technical basis, before UI and UX even come into play.
Half an hour ago I accepted an invitation to join a Team. Logged into MS in the browser, checked remember me, and then... a naked JSON response. It had a field "result", value "Success". Good! What now? Nothing else happened.
I firmly believe Teams is only usable if you're a Microsoft shop (don't even think about Firefox and friends) and is managed by a central IT (I somehow have two MS accounts under the same [work] email and no idea which is which). Don't get me started on being on Teams as an external employee.
Things like Slack just work much more and are much less surprising in their behaviour.
Totally agree with this, not to mention that chat apps are some of the most boring, uninteresting, and braindead "problems" to solve and they have been basically completely "solved" over and over again since bulletin boards.
It's different with Google products. Every new iteration of Hangouts/Chat (or whatever it's called this week) is a bit more "unsolved" than what they're killing to introduce it.
It's less boring when your tools are getting shittier over time.
Atlassian is weird in that they seem to have vacuumed up the entire market but all of their products are (AFAIK and experienced) objectively shitty and have negative demand.
A zulip clone written in end to end rust sounds interesting, with reasonable allowances for non-rust code on android/iOS. So maybe just end to end swift would also be fine.
But instead, probably instead we're going to get something like kernels written in electron forced onto us by our IT departments out of nowhere just because it's included in our Microsoft or Atlassian plan at no additional cost.
I remember at a previous company there were discussions on what to go with (HipChat/Stride had just been killed by Atlassian) and Zulip seemed to be the best free solution. But ultimately we went with Slack.
The word "enterprise" is notable in its absence in the article. Framing your problem as "enterprise xxx" would evoke Teams and HipChat, not improved Slack.
>Slack is based on IRC and is a messaging client. It was built for synchronous communication and not for collaboration. As a result, the way to scale Slack for a large company is to create more and more channels. This increases the amount of information generated in a company. Whilst this can be a good thing, Slack users don’t have the ability to filter high-quality information from low-quality. [emphasis mine]
This is the crux of the issue for me personally. The way Slack handles notifications and badges creates basically only three levels of importance: read this immediately, read this relatively quickly, or never read this. That works for small teams in which the gap between the second and third option is rather small. However as companies grow they generate more data and a huge gulf starts to form between the second and third option. There is now a wealth of information that necessitates a "you might want to read this eventually" category. Yet Slack provides no native way to handle this and the end result is that either you subject yourself to frequent interruptions for messages that aren't relevant to you or you ignore messages that are.
How do you identify if a message is of the 2.5 variety without first reading it to confirm it is not a 2? You still need to be interrupted to read the message. In my experience reminders have always been a way to manage work and not a way to manage notifications.
I misunderstood and thought you were talking about making sure you didn’t forget to fully deal with the message. I think it’s worth being notified about important messages, even if you have to deal with them later. If you’re truly too busy to triage Slack, the unread status keeps the message for you while you do non-chat work, or work out of an urgent channels synchronously and ignore the others.
>I think it’s worth being notified about important messages, even if you have to deal with them later
You are still misunderstanding. This isn't the problem. The issue is how to keep yourself from being interrupted from non-important messages. Slack provides support for messages that are not time sensitive. It encourages us to treat every channel as either worthy of some urgency to read or completely irrelevant and worthy of muting.
I guess I don’t see the problem if a channel notifies you and you either read/action immediately, or triage for later. Why is it a problem if your sender expects you to decide if you have time to deal with an important message now? If the message is long but unimportant, I’d file it in the “read quickly and move on” category. I think I’m disagreeing and not misunderstanding.
By default I turn all of my channels to Mention only. Other than that I have a few channels with low posting frequency which require urgent attention that I always have on notify. This gives me all 3 message types. I use the same pattern on Matrix and IRC as well.
One thing I think Teams does well in this space, and where I feel Slack falls short, is the organisation of channels.
With Teams you can put people in... well teams, and each team has channels and other data associated with it. This hierarchy makes it easy for people to keep a multitude of channels and data organised without actually having to do it themselves.
Another thing I have to (begrudgingly) concede to Teams is that it's essentially "free" and from the perspective of an exec it does the same thing as Slack + has better video conferencing. Even simple things like meeting with 3rd parties is very easy to do on Teams - just send an invite. Whether or not the A/V actually works during the call or peoples PC's don't run out of memory mid-meeting is another matter, but on paper it looks good.
Teams also has the BEST accessibility (at least for live captioning).
I have not used Google Meet or anything, but I'm constantly surprised by the level of sophistication from Teams live captioning and transcripts. You can be watching it "type" in real time. It'll write the wrong word, but several seconds later (as the person has added more context) you see the word, or even sentence completely change to be more on point! It's wonderful.
Google meet has a very similar feature. The missing piece is it seems to only work during the meeting when obviously it'd be great to get a downloadable and searchable transcript afterwards.
Teams makes it hard to form opinions about its potential good ideas because the totality of the user experience is so abysmal, a bit like it's hard to appreciate a details of craftmanship of a torture device from the victim POV.
We use Slack at work, and I hate it. The ratio of Signal to Noise is brutal, and there's no accountability.
But the issue is not Slack per se, Slack does what Slack does: synchronous, instant, INFORMAL communication.
If we want better processes we don't need a better Slack, we need better processes and enforcing better habits.
If we want traceability of information, we need data-rich communications and structured communications, and we have other tools to do that, from emails to databases to actual systems that reside outside Slack.
So now the problem is enforceability and not the tool per se.
My favorite solution is an Agile organization with Agile team. Vertically integrated where each lead has 4-5 reports, each responsible for projects, processes and documentation.
The problem to adoption is that the C-suite doesn't understand processes, and leads don't want to be bothered.
So the broken information system enterprise continue its crippled path, akin to walking in the dark.
The enterprise needs process innovation, not technical innovation.
I think Twist is the right answer. It's basically a forum, which is nice because:
- people have to write "real messages", with a beginning, middle, and end. Subject line.
- Doesn't suffer from e-mail's "I wasn't on the original to/cc" problem where information is completely silo'd by default
- Messages by default need to be categorized and a bit organized. Less of a firehose
E-mail gets a lot of shit, but from my experience the biggest issue is just the transmission issue. Forums/stuff like Twist solve this nicely, in my opinion.
I think Twist is totally undersold as the future. If only they would put custom emoji (this is, I swear, a blocker for moving to it at $JOB. In a full/mostly remote world, we want ways of expressing at least a bit of company culture!)
Really you want a mix of tools, of course. You want stuff for synchronous discussions, you want recorded minutes/"serious discussions", you want a place where you write up big documentation (Confluence/xWiki etc).... there's not gonna be a single tool that solves it all. And if you have shitty communication culture, all tools are going to be miserable. But there are interesting tools that go beyond "messaging" out there!
I use Twist daily and it seems abandoned. There are few UI/UX problems that hurt me constantly and I don't remember when was the last time the app was updated.
I think Zulip has the potential to become a dominant solution in this space if they find way to make the UI as appealing as Slack. Zulip offers a new approach on how to structure discussions by threads while still being a chat application with a “linear” flow.
I switched companies which meant moving from Zulip to Slack. I don’t understand why people like Slack or what they like about the UI.
Threading is terrible, async catchup is difficult, gif links don’t convert cleanly, group DMs are limited to 8 users then Slack suggests creating a new channel…?!
Zulip requires all conversations to happen inside of topics; think of it like an extremely lightweight and synchronous email, where the topic is the subject line. You can view a channel to see all its messages chronologically, and in the sidebar you can narrow to a single topic.
Zulip is great, I highly recommend it. On projects where I have to use Slack or Slack clones, I end up closing them when it's time to get work done because they're terrible at being productive and hopeless at preserving information for more than a few seconds. With Zulip I have successfully searched for information sent months ago/thousands of messages ago and still found it, thanks to the powerful search and mandatory topics.
I'm actually really interesting in this. What are the top 3 things you could change in slack / teams that would make you switch. I'll go first:
* Notifications settings are annoying I'm either swamped with notifications or miss them. I want to be able to switch of notifications for short periods
* I want to group discussions about certain things. I.e. if we are discussing a JIRA issue I can select the message and group them into something that I can attach to the issue we are discussing
* Too many private chats - I'd like to ban private chats as I find there's some great info in these which are often not shared to the group
Notifications need to be more intelligent — if Mary rarely posts in a channel (or joined the channel just to post) it might be worth more attention than John who keeps pressing enter mid stream of consciousness.
If you've a long running channel like a "developers channel" it probably has some bot integrations — "the last PR failed linting". It should be possible to mute these without muting the whole channel.
A lot of problems come down to company culture though. I think the hard work of fostering good practices is something outside of a tool's remit.
Maybe it should be a point on people's annual review if you want to get people to use it as less of a rubbish bin. "Sorry Peter, you won't get a pay rise this year as you sent an at-channel message to the team every day in Feb even when we told you that was really disruptive"(!) Most places don't care about these efficiencies enough though (yet). E.g. if people were as wasteful with company expense accounts as they are with other people's attention it would definitely get sorted out.
Yeah a lot of orgs and employees are pretty toxic and Slack and teams makes it very easy to be like that. I wonder if there is a tech solution to this or if it's something that has to be handled outside of the tool.
I'd add security to the list of things to address.
Aggregating all of the institutional knowledge of the company to one perpetually searchable global workspace hosted in a US company's shared cloud vulnerable to being downloaded and abused by a global breakin or any compromised employee account is bad.
I thought much the same. Controlling the custody and flow of information is impossible with tens of thousands of users and pretty much no policies or significant monitoring in place.
I think centralised control by company IT admins would be more catering to Teams customers than to Slack customers. To keep the attraction and user satisfaction of Slack it needs to be user / team centric. Maybe something federated or sharded like Mastodon or Discord. Might even be possible to distrust the servers and have end2end crypto.
It’s wild that Google, despite being the leader in startup email/business software and literally giving their Slack competitor to everyone for free as part of their Google Business subscription, STILL has such low mindshare in this space that it didn’t even merit a mention in the blog post.
(Full disclosure: my company used Google Chat, and I find the integration ecosystem so pitiful that I’m constantly thinking about paying up for Slack. I don’t know how Google is screwing this market up so badly.)
I think the issue (read on HN a lot of comments about the subject) is that there is no unified vision/goal inside Google, there are a lot of teams with different goals and visions, but it seems that the most common goal for their employees is to get enough merits to get promoted, forget about the old project and jump into a new one (because new things give you merits to continue climbing).
I could be wrong, with no insight information this was my conclusion after reading comments and also seen too many products killed by Google, even releasing new worst products to replace old ones that were better.
Especially because they came so close so early with Google Wave! That was it, I think, but Google didn't have the confidence at the time to stick with it, and it was just too unpolished to keep out in the public at a time when everyone thought that whatever Google did was gold.
Seriously, I think the Wave ideas were the ones that hit the perfect feature-set for informal electronic business communication.
For personal use, GChat has really been hurt ever since they merged it into Mail. It's much harder to share files and access now. Also they took away video for some reason
Eh, but you remember the first one that around when AIM was around, and pretty much nothing used today performs as good as it or AIM did on the average computer 20 years ago.
I have used Slack, Teams and Discord. If I should choose one I would go for Discord. Discord is much better in the collaboration department for distributed teams, and I love their open audio channels. It just works. Teams is good when it comes to integration with office and onedrive (of course). Slack is, in my opinion, just another chat service. A hugely successful one with a big user base and a strong brand, but not superior when it comes to collaboration.
> 2. Slack clone but with Feature X. Feature X is usually something like task management, project management, notes etc. Examples of these apps are Chanty, Rock.so, Flock, Ryver, Twist (there are many more).
> The problem with the apps in the second category is that they end up competing with two existing tools.
I use 4 written communication tools:
1. Slack for (a) async & sync team chat (b) async company/other-teams channels
2. Email for formal company communication (I don't send email, other than auto calendar invites)
3. git/GitHub for project work
4. Google Docs for RFCs
In addition there are sync meets and pairing sessions.
The only thing that I find missing is a good shared whiteboard and hallway/watercooler chats.
The problem with Slack exists between chair and keyboard--it's about finding a way to use it that works. If the culture is for anyone to post arbitrary volume in a large channel, then that's on you. Similarly if your company/teams create too many channels that many people should be in, you're doing it wrong. My policy is that if there's more than X people in a channel I don't need to pay attention (except for the 1 official company announcements channel).
In 2016, I was hoping slack or a competitor would take a native desktop application, but ultimately found slack's problem's managable.
It's a shame that instead of literally any progress or competition at all, the industry is regressing rapidly into the slower, buggier, and in every way I can think of, just all around inferior product called "teams".
I'm always curious about how different teams have "solved" whiteboarding with distributed teams, as they are normally then not whiteboarding but using a collaborative diagramming tool which is much too restrictive.
The closest I've found is making sure everyone has a decent sized tablet and using Google Jam.
Slack _is_ missing some basic functionality for keeping yourself organized when receiving a bunch of info, it's not just about the people. Especially when it's the replacement for corporate e-mail.
Dumb example: you are in N threads. You want to remove yourself from those threads. You can turn off being notified from them one-by-one (refreshing the page to actually have them disappear). Compare with e-mail, where you can quickly select a bunch of threads and just hit "archive".
I've also used Miro. It's good for gathering information, but not so good for synthesizing/creating/collaborating on a design diagram that you can do with a whiteboard and marker. Maybe it can work better with a large tablet device--I haven't tried.
>I spent nearly 4 years working on Workplace from Meta, so I’m familiar with building in this space.
Some VC guy said this is one of the biggest red flags he looks for in a startup. It almost always leads to failure. I dont remember the exact reasons but something like it means the startup doesnt understand fundamental market segment issues and grit required to overcome them.
I have a similar 'rule' (advising to small businesses) - as a company scales from 1 to 12 to 35 to 120 to 1000 people [1], they move through different phases.
So I see a lot of my clients in the 12-35 person 'Process' phase being excited when they hire someone from a 120+ person company, because of all the expertise they can bring. But that 'Strategy' phase expertise, often when they have a full team and budget to execute their ideas, rarely translates to a smaller business where they also have to roll up their sleeves and GSD.
Sitting in a team building stuff for Meta is definitely a whole other ballpark to convincing Dave's Accounting LLC to give your Slack-competitor a chance.
The statement "I worked on X at BigCo Y" does not imply that the person has enough knowledge about X to start a startup with. The problems you have when starting a new project under the wings of an existing company (especially one so large as Meta) are very different than you get when starting a new startup, so the experience does not carry over to the new company as some people expect.
Presumably making your own Slack, or more generally spending time on things you don't have a competitive advantage at by reinventing someone else's wheel
I've long held the opinion the best chat app is no chat app, i.e. delegate everything to your favorite issue tracker and other forms of knowledge keeping (e.g. Git commits, design documents).
Let everything be either fully 'literate' (for lack of a better term: abundant context, complete paragraphs and argumentation, hyperlinking), or fully sync and human (meetings), with no in between other than email, used thinly.
I'd wager that this approach will be increasingly obvious as more people burn out from messy remote settings - many of which were improvised due to the pandemic, yet are here to stay.
It's a model that has worked well for OSS - many projects do just fine with Github alone, sustainably, for years.
I work for a company distributed across the world with ~200 users in our 'watercooler' channel. I can't think of a better alternative to Slack that would enable everyone to stay in touch and feel part of a community. Email, forums, or — god forbid — enormous video chats would be overkill.
How do you feel about writing this in a community spanning thousands of people across tens of countries and providing only a fully ‘literate’ approach as per the GPs comment?
I'm familiar with the term (as popularized by http://asyncmanifesto.org/) but perhaps 'async' is not enough of a self-sufficient term. Something can be async but also happen to be poor.
I agree and there are benefits on some sync activities too.
Being better is another thing after we understand and adopt necessary concepts/strategies like 'async'. But without those the ability to improve would be more limited.
The question is loaded because while a OSS project might have chat "presence", it doesn't mean at all that its organisation revolves around said chat medium.
Chat is more often a support channel.
When you keep in mind the scarcity and distribution of OSS maintainers, the need for asynchony will be evident.
(If you want an answer: multiple 1k+ stars projects I'm a maintainer of)
It feels to me that the type of communication slack is used for can't scale to a huge company, rather than it being a tooling issue. It's useful to be able to pull disparate people into a channel when working on a project, or to be able to find someone from another team to ask a question. But, expecting people to be able to discuss things with everyone from a large organisation all of the time just seems destined for failure.
At a certain size it probably makes sense to have actual humans in each department working out what communication needs to flow to the rest of the organisation and moving it in a different tool. Maybe links to those "publications" are posted in an all hands channel that most people have muted until they're ready to sit down and catch up with what everybody else is doing, but I can't imagine a situation where they should be getting notifications in real time about it.
Not to get off topic but the problem I find (consistently) with online comms (i.e., email, Slack, Teams, FB Groups, LinkedIn, etc.) is subpar human to human communication skills.
Typically, adding (unnecessary) friction is: vagueness or ambiguity. We can't read each others' minds. We're not sitting in each other's seats / moments. My current context (i.e., what I'm focused on) has influence on where my thought process can go next. Dropping your without-warning thought bomb in the middle of that without a hint of context is too assuming. Assumption-based Comms creates unnecessary noise and friction.
And yes, these missteps at scale certainly don't help.
I think the author should invest in learning the keyboard shortcuts and search more often. They’re great and allow me to move around slack with ease. Also reminders for things you don’t want to think about at the moment work well too
The noise issue seems tool agnostic to me, if you converse with noisy people who are sporadic with posting and put important messages in threads/channels with less visibility then you’d have issues no matter what tool you use. Does anyone ever read a company wide email longer than 1 paragraph anyway? At least with slack you can search it
Matrix is even more overwhelming than slack in the same sense that email is more overwhelming than lotus notes. Maybe overwhelming is fine if the use for the tech isn't siloed to solving communication in only one part of their life.
I cant believe I read Slack and fast in one sentence.
We use mattermost and its mostly OK to the point i'm willing to use it which cannot be said about Slack. Forcing me to have Slack open all the time was reason I left a job. It was unbearable.
It's surprising to read that. No one at work hates it so much, afaik. I wonder what's different about our channel lineup, usage habits, or anything else that is such a contrast to your experience.
I think is fast if you compare with the other products mentioned, I hope they get their shit together (Slack and Teams) and build a real native app that is fast and use a lot less ram.
Some have described it as a structured Notion, but it combines lots of functionality from project management, to org charts, to threads and discussion, as well as knowledge management.
1. Super terrible at threading. Threads are clearly discouraged and don't nest. This appears to be by design, but even Usenet handled threading better.
2. Slack is an information wastebasket - at least in my experience it's really hard to find information once there's a significant traffic in a deployment.
3. Too Much Notification! I have to mute channels that has a high rate of addressing the channel.
However there's no denying that it's popular. Seems like every place I go is using Slack. At least slack allows me to edit my typos.