Challenges ahead.
A shy over a year ago, I wrote about 3 startups: Talko, Wire and Switch
All of them looked promising. All were using WebRTC.
In 2015, Switch had a meeting with $35 million, along with quite a few successful deployments in businesses big and small.
A month ago, Talko got acquired by Microsoft. I’ve interviewed here the Talko team in the past. Selling to Microsoft shows. Shutting the company. With little objections from customers. It all points to a single conclusion – Talko has been a failure when it comes to the business side of it. It probably had a solid technology – otherwise – why would Microsoft acquihire the team and fold it into Skype? I am sure Ray Ozzie and the team of Microsoft veterans in Talko added to this acquisition, but there was no other value in this transaction.
The Talko Team expresses it best on their updated homepage:
However, as engaged as many of you have been, the reality is that the broad-based success of communications apps tends to be binary: A small number of apps earn and achieve great viral growth, while most fall into some stable niche.
Talko didn’t grow fast enough or big enough. Clementine’s acquisition by Dropbox is similar. A communication solution geared towards team/group/enterprise communications gets acquired for its team with the service being left behind, never to be seen again.
And that’s in the less competitive domain of the enterprise. What will be with Wire? The third company I wrote about.
On Android, Wire reportedly has 100K-500K installs. Assuming iOS has twice as much (I am trying to be positive), that still falls way short of any of the messaging services we usually hear about – they are measured by 100’s of millions. Of active monthly users – not installs.
It is hard to see how Wire can change its abysmal future without a serious pivot or a drastic change in current market trends. Some will say this is a matter of a directory service and network effects. I think it is a matter of strategy and luck. Where Wire failed to attract the crowds, a different messaging service – Telegram, with 50M-100M installs on Android and a reported 60M monthly active users.
Wire was formed in 2012 and Telegram in 2013. So we can’t say Telegram had any head start here.
WebRTC makes it too easy to build and launch a communication service, which in turn, makes it hard to build a viable business with it. The role of product managers and people who need to think of the business case is more important than the technologists building the service when it comes to WebRTC. At the same time, finding good developers who grok WebRTC isn’t easy either.
2016 is going to be crucial for Wire.
What do you see for your initiative in 2016? Do you have a business case and route to market and money, or are you tinkering with the technology, assuming that if you build it they will come?
so where are all those product managers hiding?
I guess they still don’t know what WebRTC is capable of and where its limits are.
Your job is to inform them 🙂
So can we expect a future krankygeek event that focus on product management and business development aspects of WebRTC apps?
We probably should 🙂
Wire should clearly federate with Matrix and be a best of breed glossy messaging/VoIP UI… which isn’t Just Another Silo. I’d use it as such, but otherwise despite having it installed from day 1 i barely have reason to use it as almost all my convos happen on other networks or in Matrix. Which is a real shame, as it’s a really really nice app – just horribly constrained by the network it’s limited to. I am amazed they tried to create a new silo from scratch – just as Talko did. It simply doesn’t pay off.
How would federation with Matrix have helped Talko? I submit the opposite since any viral adoption would have slowed since most of the friends of early adopters will not be compelled to sign up for an account. So if anything, the total number of accounts or monthly users will be less than what Talko realized.
From their statement in their blog, they opted for a business model of winner take all and lost. That is that and there is nothing that can be done. It is not clear that Wire is in the same position. For some (Andy Abramson for ex) have claimed that they are playing flipping game. In which case, they know what they are doing.
The problem is Tsahi is assuming that the only way to win is to amass a large user base with a huge payoff. There are other ways of playing the game.
Aswath, while I do agree with your assesment about the uselessness of federation in growing mass audience, I have to ask – what would be that magical business model that can make a company like Wire (=consumers) viable and successful in their own eyes that doesn’t require a large user base?
I don’t think we have enough data points to know how successful federation can be for growing mass audience (which is why we should try and get some!). If done right, you can *STILL* have viral adoption whilst being able to reach out to a wider audience.
The reason people installed Wire *wasn’t* because they were forced to in order to talk to people on it (they already can talk to those people via other silos!) – but because it has a innovative UI, a polished UX, and some great unique features, and they wanted to try it out.
The way you get viral adoption is to make the differentiating features great. The basic voip/chat features then work cross-application, but the second you want to use something unique like Wire’s sketch function, or just enjoy the UI, *that*’s the point that the Wire users virally encourage their contacts to install the app. It’s not blackmailing them to join the community, instead it’s onboarding them when there’s actually a feature or functionality that *requires* them to have installed Wire. Feels pretty viral to me 🙂
why would they bet on Matrix if they could have just used a real open standard (that can survive without massive corporate sponsorship) with a much larger network than just Matrix?
How’s adoption going?
Philipp: they might bet on Matrix because it is indeed a real open standard (http://matrix.org/docs/spec) – and we’re currently incorporating a software foundation for Matrix as a limited by guarantee non-profit UK company, owned by the key contributors to the project, with no corporate strings attached whatsoever. Totally agreed that Matrix wouldn’t have got to this point without corporate sponsorship though – creating a successful comms meta-network is not trivial 🙂
They might also bet on Matrix because the featureset it provides out of the box is pretty much a direct match for a group messaging/voip app like Wire, and because Matrix has put a lot of time and focus into making bridging to existing communities relatively painless, hence the name Matrix (although we are indeed still in beta).
I haven’t been tracking how healthy XMPP’s open federated network is these days, but I’m sure Wire could also look at XMPP for extending its reach too. In terms of Matrix’s uptake – we’ve been in beta for just over a year now, and looking in the matrix.org server’s DB, I’m seeing 227 servers federated right now, 106,743 accounts, and 10,979,446 messages. We have limited visibility on metrics from other servers right now, although we did add in opt-in ‘phone home’ metrics reporting into synapse a month or so ago, which should give a bit more idea of wider adoption.
The main stuff missing on Matrix right now is a wider range of high quality bridges. We have IRC, Slack, SIP, libpurple, basic XMPP from the community and a few others – but they need much more love, and we’re fixing this currently. Also missing is a flagship glossy client (although that is about to land in the form of Vector.im); and really performant server implementation (which is in development in the form of ‘Dendron’). There are also rough bits of the spec which still need to be improved (e.g. improving the VoIP signalling); threading; editable messages; extensible profile/presence data.
The final missing piece of the puzzle is solving the problem of decentralised identity and reputation tracking to help address abuse. This can be solved entirely independently of Matrix, and would be useful for *any* open federated decentralised system – be it Matrix, XMPP, SMTP, Blockchain, etc. Is XSF looking into this? Any interest in collaborating?
If open standards are to succeed, it might be worth working together (especially when they have very different emphasis and architecture) rather than sniping in blog comment threads…
Great article Tsahi and thanks for leaving open the answer to your questions 🙂 In “Western” countries I think that on one hand users are looking for increased simplicity, on the other hand they are sticked to Google, MS and Facebook. Not much room for new players unless they start in a new region (Telegram’s founder is also the founder of VKontakte the Russian Facebook) or quickly find a core target . The answer to your question is probably to be found in the other article your wrote recently on messaging and AI. The disruptiveness of Wire, Talko and Switch is weak and it doesn’t offer any wow factor.
Luis, thanks for the comment. I’d point out that sometimes you don’t need disrupting to success – you just need a solid business plan that takes into account the changes in the market.
I think the old paradigm of “if you build it they will come” is not necessarily correct anymore, though I am not sure you need a business case from day 1.
I personally think wire can make it but not the way they are going. They have focused too much on how the app looks and not on what problem the app is solving. The only reason Messenger took off without a clear linkage to phone numbers is because it had the FB head start. All the other guys (Whatsapp, Viber, WeChat, etc, all rely on phonenumber (or FB) to get started – no usernames and passwords which are nothing more than annoying these days. Not sure I understand what the wire guys were thinking when they got out of the gate without the phone based contact list being the center of the app.
🙂 Totally agree. But you need at least to have a strong vertical or geographic to start with. I don’t know to what extent it was the case for Talko. I tend to believe more in solutions like Switch.co which might not be that disruptive but integrate in an existing working environment (ie Google for Works, MS365). And b2c is a very crowded space.