Twist – Mindful Team Communication
twistapp.com
twistapp.com
I'm the co-founder of https://www.fwdeveryone.com, which implements the same sort of functionality on top of email. This way you can capture both the knowledge created within your organization, and also conversations between your coworkers and people outside your business. There is no new client to adopt, and your data already comes pre-exported into your gmail or google apps account. And if you have a conversation you'd like to share publicly then we provide tools, like requesting permission and redacting sensitive info, to make this easy also, e.g.: https://www.reddit.com/r/Connecticut/comments/65kbok/good_hi...
Wikis are great, but maintaining them is basically a full time job, which is why they tend to be terrible and out of date. And even in cases where you have a PM maintaining your wiki full time, the majority of organizational knowledge is still locked up in email. Given that the amount of personal non-spam email sent every day is something like 100x the total size of Wikipedia, it seems unlikely that chat programs, email replacement solutions, or wikis can ever really be a complete solution.
Also, everyone already knows how to write or respond to an email, but many people within organizations don't know how to do simple things like making an account on a website. It sounds like a trivial problem, but we've found that having a system where everyone can contribute without having to learn anything new is actually a big win.
Good documentation is a full time job, via Wiki or whatever. It's not the Wiki's fault it's out of date.
Being able to use DKIM and SPF to authenticate threads is part of our value proposition, and a key part of the value proposition of email in general. (For example, being able to prove that you agreed on X project requirements with a certain person on a certain date.) Doing this type of authentication requires every message in a thread, not just the last one, so the most pragmatic way to meet this requirement is to work with email service providers with available APIs. The alternative is storing the entire contents of your inbox, but this doesn't make sense until we're ready to roll out our enterprise offering.
A threaded store and forward system works better for that as it allows new participants (members of the team) to read back in the historical documents about what happened before and what outcomes were created.
The three pillars of communication appear to me to be:
Time sensitive; communication of the specific moment (calendars are included here)
Evolved knowledge; communications of the group which represent the evolution of the group understanding of processes, challenges, and solutions.
Directive/Command; communications which direct individual or group action and recover group status.
The typical solutions fall into 'chat/calendar', 'news/bulleitin board', and 'email'
The knowledge you get at school or college is a kind "onboarding" new members. But it doesn't make you read every single piece that was ever written. You get a summary of both events and outcome, and most importantly the current state.
That's also how Wikipedia works: Articles don't get longer and longer, you can also summarize and put stuff into perspective. That's why there is a way to delete stuff (with version control, of course).
Moreover, onboarding new people (or maintaining the collective memory) isn't the only form of communication that happens in a team. Maybe it's not even the most important.
What about coordination? On the smaller stuff, on the daily stuff, but also on the bigger picture?
Now it you want to create a product/company out of that practice, then you would build a system that would let people reviewing messages be able to 'vote them up' based on their importance to the over all work product. That sort of 'auto curation' would then let you scan just the list with a 'curation threshold' and just read the important bits rather than everything in the list.
That might be a good compromise between a dedicated historian/textbook writer and a list which had all the data but it was inter spersed with the occasional "lets gets the group together for a movie / which movie" discussion.
Same if you bring new team members up to speed by talking to that person.
There is typically no single "historical" document that does the job. Old documents (e-mails, descriptions, comments, ...) tend to be outdated. New documents tend to be increments, so they don't really make sense if you don't know what is written in the old documents.
I expect that in the near future, automatic summaries based on NLP would be able to do all the heavy lifting, so the head of engineering (or whoever is in charge) will only edit the doc, not write it from scratch.
The idea that is different here is that it is an invite only system - meaning no spam.
The first question that came up for me was 'who do I invite' - I had to think who in the team is an early adopter and would be receptive to trying something new. Then my next thought was to Hangouts them and tell them what I was doing so they did not see an invite for something unknown... then I thought... Why don't I just email them... and lastly... Why am I doing this again? I have already everything I need to talk to them... What is the advantage of this service?
As it stands, I get very little spam.
So what is the value prop?
Much of what I am reading is that it has better thread handling - I will assume that it does. I have also read the part where it is better for new team members - I get that, a new member can pop in and see the trail (how often that would actually happen? - unknown). I also read the sync vs async argument - sure, I like async but at times I need sync, like hard sync with red lights and bells (sms and whatsapp + phone calls and knocking on doors at night).
I think at the core there is a kernel of greatness that can be built upon: Everything is a 'topic' and has potential triggers to external systems. I think this could be a system that has everything in it eventually. If a topic is a lazy topic then it could be tagged as such, for example a message about next weeks dev lunch, but if the site is down then tag it with a critical tag and the alarms could go off. I see the real power here is that it can be connected to external things much better than email can without significant rfc822 parsing. There is opportunity here to supplant your current email stream.
The other coolness I can see is the potential for archiving everything - I think in certain circumstances Hangouts is too ephemeral. Daily standups should be recorded for remote teams or brainstorming sessions for SR&ED (Canadian tax thing). Even whiteboards and phone calls should be archived against a topic. Connect the system to a SIP service and all calls about a topic could be logged. Perhaps WebRTC would be the cool kids way of doing things.
I would use a service that recorded everything (that I wanted to be recorded) and had ways of moving from async to hard sync fluidly.
Email is great and we are fans of it, but not for internal team communication.
> An email inbox is unorganized
> Email chains easily become long and difficult to read, and can splinter off into multiple chains
> when someone goes on vacation or leaves your team, progress grinds to a halt
> your email inbox is just one of many places where your team communicates.
[rehash of email is bad]
> On top of that, Twist gives your team the space to get deep work done by allowing you to snooze all notifications.
Email never interrupts me. I have all audible and popup notifications turned off and generally check if I have new email only when it suits me. I make this super fast and simple by checking if my email icon on my phone has a little red number. Takes two seconds. If I have email, I can read them, if I don't, great.
Like you, I use filters and labels to keep my inbox clean. Automated emails and mailing lists get filtered away right away (and I check them manually when I feel like it). My inbox itself then takes a few seconds to clean leaving only stuff I have to act on. I use the Spark mail client which is very similar to google inbox in that I can quickly deal wia h messages or "snooze" those I can't deal with right now.
This way, I have zero unread messages in my inbox most of the time. (And I get OCD when I see other peopl ea phones or screenshots with 30k unread messages... wtf is wrong with people!)
I only see a proprietary email service here, one that cannot be used with other existing solutions. Why not build this on top of the current email protocols and offer a good client for reading it and an easy way to setup a mailing list (with an easy way to browse the archives)? That way this could be used with other clients and existing mail bots. No need for another integration API or to install another bloated client. If the user experience is really what sets this apart from mailing lists then you shouldn't have problems monetizing it.
Also the point about
> An email inbox is essentially just another list to manage.
Twist is the same, but everyone already has an email account, why should I add yet another one?
The server/client mechanism is robust and scalable. With a good client UI, it could be turned into quite a smart social/work platform.
I've been toying with attempting such a NNTP client, but it's way down on my list of projects.
Of course, you can set up email group systems with archiving and search in those, but... that's more work, now you may have multiple systems for people to have access to, etc.
systems like this are logical extensions of the ceremony of email with some added benefits.
1. Must all messages be thread topics or thread replies? 2. Are messages presented in the order they were sent/recieved or in some grouping based on thread? 3. Can messages be "re-threaded" so that replies i. become their own thread topic? ii. become a reply to a different thread?
From my own research, it seems the answers are:
1. yes, all messages are either thread topics or replies; 2. messages are only grouped by thread and can't be seen as a stream where real-time communication across threads can happen; 3. and, no, messages cannot be re-threaded.
Which, if true, I think makes it decidedly worse than Flowdock. Its true that Flowdock also does not support re-threading and for that I will eternally hate Rally/CA for never adding but in all other aspects where Flowdock attempts to be something fundamentally different than email, Twist attempts to be, like Slack before it, an email killer.
For Flowdock default communication in a channel is just a message and only reifies into a thread topic when it recieves a reply. Also, the default presentation of messages is "real-time" group chat with color coding to identify threads and a dedicated thread viewer to "focus" on a single discussion if needed. Essentially Flowdock is standard group chat with an admittedly useful organizational concept of threads.
Twist seems to strip out the "group chat" part, leaving only threads and at that point I kind of agree that its just sexier email. I'd argue the whole reason Flowdock (or similiar) is necessary in addition to email is it reduces the barrier of what is "message worthy". If I want to send a message about my dog as an email, that feels weird. If I want to send the same message to a channel in Flowdock it doesn't; it may even become a thread but it doesn't disrupt useful work going on in other threads. Twist seems like it falls decidedly in the first camp and _that_ is why I won't switch to it from Flowdock. I simply can't use it for water-cooler style communication (group chat - Slack, Flowdock, even AIM win here) or productive real-time focused discussion _across problems_ (multi-threaded group chat so that one thread can dovetail/inform another in the public discussion - this is a Flowdock only thing AFAIK). The only problem it seems to solve is productive focused discussion on a single topic and honestly email already does that very well.
I will say searching is nice; of course, Flowdock has that too (as does Slack and many internet-based email solutions).
We have been using Slack as a UI for our backend services, for example, and it is a great way to present reports and alerts, and offer basic controls to the relevant person. For me, whether or not I actually adopt Twist will depend on how easily I can interact with it from my own software.
I would suggest making the integrations page easier to find - I had to troll through the Zendesk archive to find it, even though it is a core feature. Also, a link to the API docs on the help page would be nice. (Edit: Didn't realize until now that integrations are at the channel level, not the top level.)
The main reason I kept the desktop app, is so that I don't need to "download" the app every time my browser's cached copy is invalidated or expires. But being able to start the app with a couple keystrokes from my OS's quick launch menu is also nice.
I guess it's too early for me to evaluate the performance of Twist, because I am not on any large teams there. But I can see that Slack at least, has room for improvement in rendering performance. Perhaps there is some general optimization they are missing out on, like immutable state + shouldComponentUpdate. That one saved a flailing app of mine in the past.
I'm in Firefox on Linux, i7-4500U, 8GB ram.
1. Hitting "Enter" doesn't send the message. This encourages writing actual paragraphs in response, rather than sending fragments of sentences
that can be interrupted across multiple lines
because you're not really sure if you've finished your sentence just yet
oh and you thought of one more thing so everyone else please take this line into consideration as well
Of course, you might say "train your team to not do that." Sure. But I'd rather use a tool that doesn't require breaking (perhaps reasonable) habits. And if communication shouldn't be through short bursty chunks by default, why make that the easiest thing to do?
2. Make threads more of a default way to respond, and make thread comments first-class citizens. Presented with the UX of Slack, it's really hard to move yourself towards using threads because the easiest way to respond to things is to type and hit enter. Also, threads live in their own "All threads" space, not organized by channel. Thread comments aren't first-class citizens like regular chat is because they can't have files attached, be posts or snippets, etc, so sometimes it feels like I actually _lose_ functionality by starting a thread.
Twist looks like it has what I love about Slack – good for newbies to join in and see organized, curated history (I can't show new folks _my_ inbox labels and organization, or spin up a new channel on a mailing list for each topic we want to discuss), good for lurking on projects that aren't your own but are related, and has emoji responses for celebration, quick feedback, and commiseration.
That said, in 3 years at Slack-heavy companies, the level of interleaved conversation I've encountered is almost nil. Maybe that's cultural (though the consistency across multiple companies and teams suggests not). Maybe company or team size is a factor. Maybe it's a matter of breaking it out into enough channels that if there are multiple topics to discuss at once they're usually happening in separate channels anyway.
Sure, Slack could probably improve the threads UI. (Though I happen to think that threaded display and real-time discussion are in some ways conflicting goals.) As it stands, however, they're a black hole. My team has actually banned the use of threads (and no, it wasn't even my suggestion).
And it's good.
I think internet went wrong way to ditch traditional internet forums/discussion groups/ and internet communication started to be based on real time chats (Slack, Messenger, Skype etc.), but I'm glad that we're returning to the oldschool (I'm waiting for Facebook to introduce "threads/topics" on its groups, because now Facebook groups are less functional than plain old phpBB.
[ Problem ]
- Meetings have the potential to really keep a team together
- Teams are distributed throughout geography, time zones or daily life patterns
- So how to bring GOOD meetings online, and perhaps even make them asynchronous?
"Meetings" come in different flavors, for example
- Weekly or monthly meetings to look back, learn, and discuss how to move on
- Daily scrum / daily standup to get a shared view of the day
- Ad hoc meetings to address a smaller but immediate issue
- Water cooler meetings which allow for serendipity
...
So for meetings in person, the real challenge is to get everyone attending, even if there is lot of urgent / important work to do. On the flip side: Keep the signal / noise ratio of the meeting so everyone feels it's time well spent.
[ Solution ]
So my vision for a better tool would include:
- Separation of reading mode and writing mode. You should be able to write in the sense of offloading thoughts without the need / option to read through everyone else's notes at the same time
- Asynchronous default. Stuff you write should be considered a draft until intentionally "published" (exception: urgent stuff / water cooler stuff). That would allow revising bigger thoughts before a "meeting".
- Even asynchronous, virtual meetings should have schedules. So for "daily" meetings, there should be a function that makes sure that everyone reads and writes at least once a day. For "monthly" at least once a month, but without that becoming a continuous "monthly meeting", which wouldn't allow to really focus on it.
And we also have a very powerful API: https://api.twistapp.com/
And we are working on a lot more integrations.
What were the issues that Google Wave had? Why didn't it succeed?
Google was unable to properly explain to people what wave did and why they should use it, then killed it after 2 years citing adoption numbers.
I know they spent a couple years getting to this point, but I can really appreciate the quick, well-designed iOS app, and the native, non-Chromium / non-Electron Mac app that only uses 35MB of RAM and has no noticeable idle CPU draw. It's such a stark contrast with Slack and most of today's pseudo-native apps.
Impressive launch!
It's more a mix of wiki+email+chat.
Most of my friend were unable to understand the UI. The thing was sloooooow. Trying to do anything meaningful would require a lot of tinkering.
But Twists API to create either thread/comment in an integration will be more complicated to do than Slack and their attachments are not very powerful (yet?), migrating our existing Slack related code might proof difficult.
Why not just use existing infrastructure of emails and integrate with that nicely? The channels on Twist could be folders on gmails (with google integration at first) and users could reply to threads from their favorite email client. If they want to use a dedicated app they could easily install the app.
I think this would help them with growth a lot more since it would use the existing infrastructure of emails. Of course, it doesn't solve the notifications problem that they are trying to solve but that's what the app is for.
it looks a bit more complicated than the Slack one, because of the thread/comment difference, but the "Automatic URL Report" and "Loop in" email based input looks very powerful as well.
One thing that would make me consider it, if I could open multiple windows at the same time for example.