Is Slack Really Worth $2.8B? A Conversation with Stewart Butterfield
bits.blogs.nytimes.com
bits.blogs.nytimes.com
No one in technology should ever find themselves uttering these words. It's like asking for it.
If the tech world remains relatively unchanged Slack is easily worth $3 billion. All they need to do is get a few million workers on there chatting away and they've got ~$10/month * millions, which is hundreds of millions per year.
They should probably sell to Microsoft and let them take on the risk at this point. There's a good chance the tech world will change around them and the whole thing will come tumbling back down to earth. Or they could try to become the new Microsoft.
It's a completely different matter to be saying the same words in private vs in a public interview that will be read by millions of readers. A CEO must say this (or something to the general effect while being PC) when speaking in public. Doing otherwise can only have negative consequences with respect to employees, customers, etc.
But Microsoft is the most likely buyer, it's been on a spree with integrating best-in-class enterprise products/productivity tools into Office 365 (Sunrise, Accompli, previously Yammer). Slack fits into that mold, though the sale price would probably be orders of magnitude higher than those acquisitions (likely upwards of $10 billion, possibly much more). At that price tag, it might be untenable.
Not sure what Slack's medium-longer terms goals are, honestly. Their valuation is so high that only megacorps can afford them, but there's still enough risk that public markets might be skeptical (their revenue growth is astronomical, but they haven't been generating revenue for THAT long). In the interview, Stewart mentioned market saturation eventually limiting their crazy growth numbers, but I'd imagine they have a ways to go, given that the total enterprise market has to be well more than 200,000 total seats. Exciting times for them.
To what extend that's literally true or it's just that he's just working with some of the same people and tools, I don't know, but it would certainly make sense.
In much the same way Slack rose from the ashes of Glitch, Flickr did from Game Neverending, the predecessor to Glitch.
http://gaming.wikia.com/wiki/Game_Neverending
>Although development of the game was later shutdown in 2004, the tools built for GNE later evolved into Flickr, a widely-hailed photo-sharing service. Occasional signs of this legacy are visible, such as the '.gne' file extension appearing in Flickr's URLs.
http://techcrunch.com/2012/11/15/as-flickr-co-founder-butter...
>Whether or not the next venture is based around imaging services, it looks like it will have a social element: “We have developed some unique messaging technology with applications outside of the gaming world and a smaller core team will be working to develop new products,” he writes.
200k users could be achieved by signing up 3 or 4 megacorps (obviously at enterprise rates far lower than $10 per user) and having a business on that scale that doesn't make you a unicorn, even if you have a decent product and profitable business model. In fairness to Slack, the revenue run rates not mentioned in the article are more impressive, though I haven't seen how much they're spending to acquire those users.
Skype had well over 50x the user base of Slack - achieved over a similar timescale - when they sold for a similar valuation back in 2005. Frankly, I was far more impressed with how circumspect Butterfield was about his startup's prospects than the growth figures.
Microsoft would want to buy Slack for the same reason they spent $$$ on best-in-class products like Yammer, Sunrise, rebranding Acompli as Outlook. Slack's revenue is incredible for a startup, but a drop in the bucket for MSFT (for now). More immediately, Slack represents an entry point for Microsoft to upsell Slack's customers to other Office365 services and Azure services.
What supports this claim? Yes tech and the tech world changes quickly but some of the core functions of running a business, such as effective team communication, will not go away any time soon.
It's a pity that Atlassian is still napping with Hipchat usability features (particularly multiple accounts).
Communication itself won't go away, but that doesn't imply that a platform/application which facilitates communication won't fade away. Just ask IRC, ICQ and AIM (All of which are still in active use today)
It doesn't matter if the demand doesn't go away. A higher quality supply will displace Slack, just as Slack displaced the many before it.
There are several reasons why they are doing well and they have much less to do with execution than with the right people getting involved.
Then we tried Slack, and it's the first time the entire team has adopted 100% and maintained it. Slack mobile has even replaced text messaging/GroupMe for a lot of us as well.
Don't hand-wave away their success by giving credit to investors. It makes you sound bitter and insecure, or at least uninformed. They've done a great job building their product and deserve the position they're in. Whether or not $2.8B is an accurate valuation is beyond me, but it does seem like they're minting money over there.
But it's not just an execution game as many would like to claim.
Nothing prevents anyone else from executing better. The permanence of the need doesn't point, in any way, to the permanence of this particular supply.
Using multiple devices became mainstream; nowadays everyone needs to be able to use your system from their phone and their computer at the same time. This shouldn't be hard to handle but through accidents of history it's terrible on the big open-source options (Jabber or IRC). And companies need internally-hosted. If AOL had been willing to make a self-hosted AIM, or MS a self-hosted Skype, they would've killed it.
I'm not sure why they didn't. And I'm not sure why Atlassian didn't seize the whole market with HipChat, which is almost identical to Slack (it's very marginally less polished). At my current company we use Slack's free tier because we fit within the usage limits for that but not those for HipChat, so maybe it's that?
But tools come and go my friend
We may go back to Slack at some point, if we can get it to work effectively for us.
The assumption is that if you're actively in the channel, you saw the item already and don't need a duplication notification on your phone for the item you just read.
I stopped using the desktop client for this reason and now I have a separate window open for the web client. Push works much better for me now.
The v2 Android app is currently in beta - I've been testing it for several weeks. It's lightyears ahead of the current release version which is admittedly pretty faulty (no caching means that resumes are very slow). I've also found notifications to be more reliable. Perhaps give it another try once they ship the v2 Android app.
But then, I thought the same thing about Twitter (how is this not worse that every other IRC / IM service I've ever used?) and look what happened there.
Not to mention the integrations outside of developers. I haven't worked in enough large companies, and never as a nontechnical employee so I'm not sure what sales teams were using for communication but I doubt IRC was their cup of tea. I imagine they used a lot of email, maybe some CRM communication but that's still not as useful, or rather I imagine it's less useful than Slack. Having a single place to discuss Stripe purchases, hiring with Breezy, MailChimp et cetera that isn't a gigantic email forward sounds seriously useful and I don't know that anything like that exists on IRC.
There are alternatives like HipChat, but they clearly didn't it the mark the way Slack has.
What Slack does very well is giving a good enough experience for programmers to do shop-talk, and ample opportunities for chatbot yak-shaves, it's also perfectly well usable for non technical people. I work for $ENORMOCORP, and after a couple of brave pioneers among the tech people, now everyone's falling over themselves to get their teams a Slack instance. (As others have said here, the ability to have multiple instances is pretty fundamental to this dynamic - we have team-specific Slacks, product specific, project specific and what have you. I certainly would not want some more 'old fashioned' colleagues seeing that NSFW bollocks people keep wiring into the tech specific chatbots...)
This strikes me as a pattern with some potential for growth - tech people driving adoption among non-technical colleagues. I reserve judgement on whether that's $3bn worth of potential.
I think the list will be much shorter for Slack compared to its peers.
The actual costs / complexity involved in building a similar product are magnitudes less than the high valuation would suggest.
So I'm not sure how much early success is really worth in this market.
Ui knowledge
Inertia
History.
The last is most important--if cultural knowledge over years gets embedded in slack, it will be easy to justify the modest ongoing costs.
Nothing excites me more than having to think about supporting yet another codebase deployed to non-technical people who will undoubtedly cause more bugs than thought possible and open more JIRA tickets about it with increasingly vague title names.
At some point Slack or [integration of choice] is cheaper than having an internal engineering team to write half-baked communication solutions.
1. A company-wide and approved solution (proprietary) that's very buggy.
2. A jabber server someone threw together so everyone could use their client or choice.
3. Google chat (through various clients).
4. MSN messenger / Skype.
And they all get used by different people for different things. It's very frustrating. Switching to a new thing would be a non-starter, probably (and we couldn't move to Slack anyway, but that's another issue).
The fact that they already have a popular competitor in the space (hipchat) would normally make such a high valuation seemingly more confusing, but if viewed using the force multiplier logic, maybe less so? There could be concern that having only a single frequently used tool would make it too easy for someone to "own" a startup's lifeblood, so better to have more than one. (similar to github and bitbucket?)
I guess the alternative is that slack has some investor specific info about their product pipline where they are pitching other companion tools that work well together, and/or compete more directly with other popular services. Like atlassian/jira or something.
I've been using Slack in a pretty casual way recently as my company switched to it. It seems, you know... fine. Like it has the same features that all the other messaging products have. What's the compelling feature of Slack?
I had a "this is what all the fuss is about?" reaction.
I actually seem to prefer the hipchat UI personally, and apparently so do many of my colleagues. The fact that hipchat is cheaper only made it clearer to us.
note: only using and comparing the osx desktop app for both services.
One key example is the multiple teams feature: the ability to have multiple accounts logged in at the same time, and the ability to easily swipe between them on iOS or toggle on desktop is a huge advantage.
A quick look at HipChat indicates this still isn't possible [1].
Considering more and more people wear different startup / company / organisation hats not having this feature is a great way to limit participation / growth in a communications platform.
That Slack had this feature at the outset for me meant I've already spread use of Slack to multiple orgs and participate in all of them at the same time.
[1] https://help.hipchat.com/forums/138883-suggestions/suggestio...
Nonsense. Slack is very beautiful with everything around the core feature, but the core feature is awful. The main chat window wastes incredible amounts of space, even in the compressed mode. The least visible items on the screen (apart from timestamps) are the chat messages from users - instead their names, icons, channel names, navigation elements are HUGE. Cute niceties like emotes are screwy due to the standard layout. We're using the Jira integration, and it wastes so much space that one entire page is taken up by a mere 3 Jira ticket updates. Stick it into 'compressed' mode and now every chat line starts at a different point, depending on the user name. Hipchat has none of the problems I've listed here, though it has it's own - generally they're issues outside the core feature.
For people used to chat elsewhere, these issues are pretty obvious. But Slack seems to be marketing to people who've never used chat, so they get to set their expectations from scratch.
Well obviously we disagree on this point, as to me something being beautiful necessarily has to encompass the UX, and I think the Slack UX is great.
As a previous long-time Jira / Confluence user I also know about atrocious UX, and Atlassian delivers that in spades (along with appallingly slow page refreshes). If you've been sensitised to the kind of UX Atlassian delivers I'm really surprised you take issue with Slack, but each to their own.
> For people used to chat elsewhere, these issues are pretty obvious. But Slack seems to be marketing to people who've never used chat, so they get to set their expectations from scratch.
Our team are long-time IRC users and have used a lot of comms solutions over the years, and Slack is the first decent one we've come across.
So I'd obviously refute your assertion that savvy users dislike Slack and unsophisticated users like Slack.
Our team are also longterm IRC users (I get mocked for only being in half-a-dozen channels) and our Slack experiment lasted only one afternoon before everyone abandoned it. The only dev who was really fighting for Slack was wanting to use their IRC integration and bypass the UI completely.
So I'd obviously refute your assertion that savvy users dislike Slack and unsophisticated users like Slack.
Refute away, because that's not what I said. Your entire rebuttal to what I said has been a strawman, apart from 'we disagree'. I gave you some very concrete examples of my problems with the Slack interface, and you give me a vague dismissal about entirely different products.
Not currently possible, but officially confirmed as being worked on.
"It’s official: Development is underway for multiple account support in the HipChat clients. Stay tuned because updates are just around the corner."
But if you're right, it's no good and there's no switching costs, then I guess they'll die. Right?
It's also 3.5x more expensive. In absolute terms that's not much (<$400 a month), but no one could point to $400 of incremental value in it. If we fully embraced it and turned off emails for those teams, would we see it? Maybe, but we'd still need emails to deal with external vendors, and there's no way that we'd get non-developers to see it as an email replacement.
I find it amusing whenever I drive by their Office Space bill board on the freeway and think to myself, "How many times will I have to send that memo through HipChat before the message gets through?"
Slack is one of the first companies walking around with a billion dollar valuation that makes me go... "Yeah, maybe..." We use it. We love it. It's mature, well designed, gets improved constantly, and is used by more local businesses I know than not. It provides real, actual benefit to our small organization above and beyond what we pay for it.
Slack, to me, is a real product, with a real revenue stream and real value to teams. I really hope they never get bought and acquired or rolled in to a larger organization. The app is a constant joy to use - I can only say that about one other program I use daily.
For me, there is also a bunch of minor detail where Slack wins. Slack just looks better, behaves better with choppy connections, and has better code pasting support. It also does much better with automatic link extraction (HipChat's Linky is annoying and won't let you remove the automatic description). Lastly, Slack has integration playbooks that automatically uses APIs to configure stuff or (when that's not possible) gives you the recipe; for example, the Github and Zendesk integrations will automatically register the app and webhooks for you, almost no manual setup needed. HipChat is all manual.
And for the love of God, HipChat, fix your inane emojis. Immature 4Chan memes (trollface etc.) have no place in a business, not even among developers.
Oh, they do. E.g. to make fun of those trying to look sooo professional.
[1] http://old.ycombinator.com/ideas.html (pg, 2008, see number 28)
Ah, because I would only ever want to talk to people within my specific team...
People need to understand that investment is a vote of confidence in the team, not just the product. What can this team accomplish with that investment? That's the question.
1. Bubble when I get a direct message. 2. Leave all but about 3-4 groups. 3. Tell images to not default display (have to click on them) 4. Star important channels (The star is next to the #channelname at the top and it's only visible when hovered) That big star on the top right is not to star a channel.
5. and probably most importantly was to switch to the compressed mode so all the chat spaces are as tightly compacted as possible.
Hipchat gets notifications right. Slack and Hall seem to have (in)sane defaults, and not quite right customizations.
The web-view of Slack is slow and stupid.
It's picky, but I loved the vim/sed style substitutions in Hipchat. Slack FTL.
But the biggest annoyance by far: In Hipchat you could reorder your chats however you felt like. In Slack they're fixed. And team chats are arbitrarily (feeling) broken up into groups and channels. Which is a completely useless distinction for most I think.
Hipchat had @all and @here. Which seem pretty self explanatory. Slack has @group, @channel, and... I guess that's it. No version of @here AFAICT, and depending on what sort of "room" you're in, the @all equivalent changes.
The iOS app also makes something that looks like a room picker and instead makes it some other menu I forget. And notifications will happily occur on my laptop, computer, iPhone and iPad all at once if I don't catch it at my desk in time. Here's a hint: If you decide to notify my phone, don't notify anything else at that point. No cat picture or CI build notification deserves Def-Con 5 treatment.
But it has themes? And useless giphy integrations? Honestly I don't get the love at all. Form over function at it's worst. I really dislike it. :-p
This seems to come about because any HipChat message to a room you are in notifies & alerts the whole room on OS X. Because of this, most communication happens in private person to person rooms which defeats the purpose of having public rooms for transparency. I haven't been able to get any of the rooms feature to stick and so what commonly happens is two people discuss things on chat, then repeat those conversations to everyone who should have been on it in meatspace.
All of this because HipChat still doesn't let you set per-room notifications.
Otherwise there'd be no point of @all and @here if you were going to be notified anyways.
Put the 'SLACK' user on ignore and most of the annoying integrations disappear.
That said, I still hate the thing.
also this way you own the data/service for the most part.
Soooo time to create a Slack clone and sell it to Slack for a few million?
Does it import/export to Excel?
(And go use Flowdock.)
No no no. Those Aerons go for $50 each on Craigslist.
Only if you can't cash out at the higher price first :), but joking aside I think your point is an important one that gets overlooked a lot when people talk about tech companies.