Hacker News new | past | comments | ask | show | jobs | submit login
Messaging apps shouldn't make money – Pavel Durov and Telegram (wired.co.uk)
146 points by bndr on Mar 29, 2015 | hide | past | web | favorite | 68 comments

I think the key thing here is that Durov has the money from selling stock in his previous company (12% w/valuation of 3-4 billion) which means he can afford to pay to support the app in the meantime without any external financial help. I do think there are different ways to monetize messaging outside of pushing advertisements into people's face or selling information and third-party paid apps are a good option.

In that regard I think the title is misleading: Durov plans to make money through the app eventually, just not through selling people's information or push advertising.

Nobody sets out to do evil but until he creates an ownership structure which prevents himself from cashing in once he's won everyone's trust this is just an underdog doing whatever it takes to attract customers.

To be succinct, one day he will sell out and the buyers will have every incentive to make money out of their users.

What's funny is that he was also promising to not put any ads at VK (russian social network he founded) ever, but pretty soon he acted in an opposite way.

I agree with everything you said and I think protecting users versus monetizing is a fine line that every field has to work with.

You can always say that the messaging app itself isn't making money but the underlying technology is.

I think that is a good point. For start-ups who are interested in making messaging services, I wonder what Durov's recommendation would be as far as financials go. One has to be able to support the team they are working with and there are few people out there who can take the route Durov has due to his financial freedom.

In certain way, cost of doing business has reduced because of cloud service providers. However, network effects and the scale requirements still add up and make it hard for startup's to work on their long term vision. It would be nice if computation was similar to electricity. That way, the onus of paying for the services lies on the end users - the more you use, the more you pay. It is more straightforward and startups can sustain paying themselves for few years without having to worry about paying for computation, storage and network.

No such thing as a free lunch.

Telegram hasn't done a very good job of eliminating the possibility of profit. The secret chat feature is actually a marginal use case. By default messages are readable by the back end. In fact, group chats are also reasable by the back end always, and any conversation which can be accessed from multiple devices. And last I heard, the official windows desktop app does not have secret chats.

All of these messages are the normal use case on telegram. And the text is being indexed by the back end for easy searching capabilities. So essentially telegram has built a search engine for the conversations of tens of millions of people who are slightly more paranoid and want a secure messenger. Not only that, many of the tens of millions have shared their entire address book with names and phone numbers because the official clients required them to.

If they're trying to avoid having valuable information that someone would pay money for, they're not doing a very good job.

This read like an ad for Telegram.

Yeah, but everything's true. Encrypted by default, huge bounty for security researches, fast, great cross platform apps.

Encrypted with their home-grown crypto. Telegram is not fully free and open source. Encryption by default? That has to be new, it was not when I recently checked. And of course there is http://www.cryptofails.com/post/70546720222/telegrams-crypta...

Use Signal/Textsecure, folks.

> Encrypted by default

However, it does not use the end to end encryption "by default".

Relevant: https://www.haystacksoftware.com/blog/2014/10/its-not-about-...

Since the early 90's, newcomers to the Internet have attempted to monetize the existing phenomenon which the Internet - in all its true free glory - has allowed.

There have been countless attempts to 'own email', and before that, 'own Gopher' and 'own Archie' and so on. There was, once, Hotmail. Then came GMail, and then .. Facebook.

It's quite possible to maintain your own social network with email alone. You can achieve the same results with a mailing list as you can with the Facebook timeline.

The only difference is that you must educate the people in your network to understand how the technology works, and how to operate it in order to function in as effective a manner as is required to maintain a high standard of communication. The fact that this is nearly impossible in the modern context - that you cannot get a group of 200+ people to operate email in the same, standardized manner, is what Facebook - and other mass-communication media - depend upon in order to gain their market share.

I would abandon Facebook in 5 minutes if I could be guaranteed that the people I care about - the links in my social network - were as competent at handling the solution as would be required in order to maintain the quality of communication that Facebook currently delivers. Alas, this is not possible - and those who are attempting to monetize and profit from this mass stupidity will do little to address the issue - for after all, it matters that people don't know how to write a proper email. This is the only reason Facebook has survived as long as it has.

If you want to truly revolutionize communications in this day and age, do something that allows people to use the existing, free, technologies of communication to increase the quality of communication among the masses. Alas, this is a very difficult problem - and like all social ills, the solution is education. The absolute hardest problem there is to solve, out there, today: teaching someone how to use an existing technology to effectively communicate.

Woah. Hardcore encryption? Is that even better than military grade encryption?

I don't think any reasonable person would refer to Telegram's horrific chimera as "hardcore encryption."

I wish there was something like Bitcoin for messaging. Users need a better way to prove their identities (or hide them, depending on the situation) and be able to verify peers. The current hierarchical security system is probably destined to require money, but the free alternatives make me uncomfortable because they tend to need private keys.

What I really want is an open version of Skype where the private key can always be generated by the user's mind, overcomes all the limitations like NAT that make users second class netizens, has no central tracker, and (optionally, but possibly required to work around the limitations) allows a certain amount of traffic to be forwarded like Tor.

I don't see how we can have a conversation about who pays for messaging when the underlying infrastructure for truly secure online communication hasn't yet gone mainstream. I don't really trust Freenet or Tor and can't quite figure out how come, so maybe we could start there and figure out why that is.

And a billion dollars for anything messaging-related completely blows my mind. Most of the world-changing software was developed for a few thousand dollars by a couple of people working in their spare time. To me, paying that much completely misses the point and probably hinders progress by locking capital up in niches that exclude developers who are working towards creating a non-scarcity based economy.

I have developed an instant messenger for Android called EnsiChat [1].

For identity, I use an RSA key pair. For key exchange, devices send their public key to direct neighbors, and you can visually check that the key is correct.

I have also thought about implementing transport over internet (right now it's only Bluetooth). But that would require a lot of effort.

[1] https://github.com/Nutomic/ensichat

http://Matrix.org makes e2e pgp encryption of audio/video/text chat friendly and federated messaging.

Bitcoin is on their radar for identity proof, as one of many ways to verify https://matrix.org/docs/spec/#n-factor-authentication and could use your help

In 5 years time a widespread conclusion might be that Telegram is the Bitcoin of messaging.

(But probably not in a way where anybody is glad about it)

That's really cool, I like how you just give the other user a Tox id token and it sets up the secure channel. I wonder if that's a general approach for secure communication, because as long as it's sent through another secure channel (that may use the traditional SSL certificates from an authority) it could be used to avoid a MITM attack.

Maybe a way to do this in the future with an insecure channel is to give the peer your contact token using only context that both of you know. So the above token could get converted to a human-readable string like "that place we met that first time and what you said to me with 12345abcdef on the end" and the other party would have to know you to figure out how to find you.

Proving who you are beyond that point would probably involve some back and forth questioning like in spy movies or a video call (which could be faked in 5-10 years with computer rendering) so shared history will be key.

It already sounds like it can run over Tor, but it would be nice if it had a standard socket wrapper that could be run over UDP NAT tunneling to look like an SSH VPN. Maybe WebRTC could be integrated with it somehow. I just mean that I'd like the negotiation phase to be separate so that if it works, you're just presented with a secure socket.

Just the opposite. The free software movement has led to apps and programs where the users are the product, not the customer, and companies bend over backwards for advertisers. The companies have no incentive to care about your privacy, just your data.

Paid messaging apps, however (at least until they get acquired by advertising companies like Facebook), give companies incentives to service their customers first, and the users are actually the customers. That's why Slack and WhatsApp and HipChat and Basecamp have taken off--because they offer the security and paid customer/company relationship.

This article's title is woefully non-conclusive. Sure it would be wonderful if ALL messaging apps could by built and funded by Russian billionaires such they were private and free for the end user, but the fact is that to support the infrastructure of a billion+ user messaging app, you need to make money. And if you can't make money from the users to whom you are promising security and privacy (and thus to whom you owe allegiance), you have to make money elsewhere. And the alternative is arguably what has caused so many security and privacy problems in the first place.

> The free software movement has led to apps and programs where the users are the product, not the customer, and companies bend over backwards for advertisers.

What part of "Shouldn't make money" didn't you understand? You do realize that if you're selling user data that makes money?

The free software movement has done no such thing. The apps and programs you're talking about where the users are the product are free as in beer, not free as in freedom, and the latter is what the free software movement is about.

> This article's title is woefully non-conclusive.

Maybe try reading more than the title?

It seems you missed this chunk of the article:

<quote>Will he continue to bankroll the business -- or does he see revenue opportunities? "We will become financially sustainable at some point," he says. "It will most likely involve third-party paid apps built on the Telegram platform."</quote>

How do you collect rent on an open platform?

"How do you make money off search?" -- question by a lot of smart people around 2000.

The commenter above, bmmayer1, noted advertisers would be how that would be done. I believe a similar method was used with search engines.

> Just the opposite. The free software movement has led to apps and programs where the users are the product, not the customer, and companies bend over backwards for advertisers.

That's a business model common for SAAS applications, I don't really think it's related to free software. Gcc on linux doesn't send usage statistics to someone, I hope.

You're right. I should have said consumer software, not free software.

Messaging services (with say >100k users) should fall under general laws for telecommunications. For instance, this would solve the problem that messaging services don't interface with eachother (lock-in), and it would solve the problem of the provider peeking into any of the messages for whatever reason.

Couldn't disagree more. Regulation has not solved any of these concerns, in fact quite the opposite. Regulation of telecommunication has resulted in some of the biggest and well-seated monopolies on the planet. Please, god, no. Messaging services should fall under exactly one law of the land, and that's the First Amendment.

But what if I want a messaging service to specifically be disconnected with other services?

By your offer, not only that you are forcing me to connect with everyone, you take an essential part of my privacy. Your suggestion is anti-freedom by its core.

Have been using Telegram for work for a few months now instead of (or in addition to) IRC - it's been very good, the mobile and desktop clients are great.

Tested Tox last year but it was super buggy. TextSecure looked good but doesn't seem to have as many users. I think Telegram will keep growing.

I just watched Durov's talk with Jimmy Wales from a couple of years back and he comes across like he genuinely cares about ideas that benefit all internet users more so than money. https://www.youtube.com/watch?v=hEHd4HbOLYM

Still, doesn't seem like a sustainable path to have the messaging infrastructure be a responsibility of one company, both for security and financial reasons. Surely some kind of tox style p2p infrastructure would be better?

Interestingly in the talk he showed that VK had 33 million visitors monthly and Telegram apparently already has 35 million.

His new project is even more successful.

> "We're trying to create a new type of IT company, one that never focuses on maximising profits, but instead provides value to society,"

This is dreadful. The Free software movement has existed for decades and Telegram brings nothing new to the table that Google hadn't already a decade ago (open yet crippled protocol, proprietary servers). There is nothing new here, besides the fact that Telegram basically makes no money. Which is also nothing new, since Google Talk and Voice never made money either!

I don't mind it, because we know the protocol and can reengineer the clients easily in the event they betray their users trust. It is much better than WhatsApp or Skype being a whole-pipe black box. But at the end of the day, XMPP has existed, has improved, and today provides all the features of all these other messaging transports and more, the problem is there is no money to be had in providing an easy signup XMPP service, so nobody advertises them. Which makes Telegram perplexing, because their only valuable resource is their userbase, and they are not directly monetizing it at all. Contrast that with Duckduckgo having an XMPP server @dukgo.com that basically got a blog post and hasn't been talked about much since, that I use all the time on all my devices.

But fundamentally none of this absolves how stupid all these companies re-implement a wheel poorly (XMPP) because the old wheel is so well known and understood that you can't lock people in or profit off the users using it.

XMPP have troubles with synchronizing chats between clients, with push notifications, with user-friendly mobile clients. I like XMPP, I did set up my own server, but I finally abandoned it. It's usable only with one client, preferably on desktop.

>synchronizing chats between clients,

The Carbons XEP[0] is now fairly widely supported and does exactly this.

>user-friendly mobile clients.

Conversations[1] is excellent and is free software.

>push notifications

The centralized system of push notifactions on Android and iOS is not yet supported, it's true, but XMPP with its extensions has pretty good support for battery-constrained devices with intermittent network connections.

I use XMPP across multiple devices with my own server to communicate to other people running their own server and it works great. Setting up your own server is extremely easy ("apt-get install prosody" works out of the box on Debian; stay away from ejabberd which is a pain) and I highly recommend it.

[0] http://xmpp.org/extensions/xep-0280.html

[1] https://play.google.com/store/apps/details?id=eu.siacs.conve...

> >synchronizing chats between clients,

> The Carbons XEP is now fairly widely supported and does exactly this

I'll add MAM (http://xmpp.org/extensions/xep-0313.html) for messages you didn't receive because one of your resources was offline, which Conversations does, along with at least Gajim.

> The centralized system of push notifactions on Android and iOS is not yet supported

It's coming (http://www.xmpp.org/extensions/xep-0357.html)

Thanks, I'll look into that. I'm more interested with iOS and OS X apps (though Pidgin works well for OS X). And iOS doesn't permit messaging apps to work in background as far as I know, so push notifications are needed for comfort work. Though I'm usually near my laptop and probably can live without them.

Not to diminish the good sides, but multi-device XMPP experience is a quite complicated matter even today. For example, you'll have to give up on OTR, and you'll have to choose the software carefully.

You don't really have to give up on OTR, but OTR has to get a lot more user-friendly for that use case. Right now it breaks all over the place, and with some really obvious bad UX, too.

This raises the obvious question, then who is going to pay for them?

Who pays for e-mail?

Who pays for e-mail?

This is the question of a zen master. In four words you've said more than the entire hagiography of TFA. Yes of course, someone has to pay for electricity, but when using an open protocol the user chooses who pays and what they're paying for. If Facebook could go back in time and kill email, they would in a heartbeat. The communications of free people should not be supervised, whether by dictators or by capitalists.

E-mail is an open protocol, but clients and hosting are generally not free. You can pay for them, or you can subscribe to ad-supported services where they data-mine your inbox so they can target ads to you. But someone is paying for it.

Interesting question. Anyone and everyone.

I may host my own email server or I might pay someone to host it. Or the government could host it, or my university could host it. Or I could get it for free by agreeing to be served some ads.

But who pays for it doesn't matter. I can send you an email even if we use very different providers (from the point of view of both technology and payment). For messaging I don't see this yet.

I feel like XMPP was supposed to be the chosen one. I had a google account but chatted to my gmail contacts through my self-hosted jabber server. I wonder if this new generation of messaging apps will get to that point anytime soon.

Private companies big enough to warrant a proper GApps account or their own mail server (or many).

In gmail - ads

Google and every other company that invests the immense time and money required to keep it clear of spam. Email without a large spam filtering infrastructure is unusable.

I used email before there were infrastructure-based spam filters. It was very usable with client-based spam filters. You can't do encrypted email without either a whitelisting system or client-based spam filters.

With my personal email - google. They handle the costs of storage and transport and in exchange they show me ads and collect high level data. With my work email - my company. They pay to keep the email secure and contextual and in exchange I keep it mostly related to work.

The cost of storing and transporting e-mail is potentially so low that I would claim you are overpaying Google.

users could certainly pay ..but through a donation model similar to public radio or maybe Wikipedia. there's a carefull line to walk if 'sponsors' come on board but it's possible.

Is it open source and open protocol?

If so, that's what we need. Email was once a bunch of proprietary systems - MCIMail, CompuServ, AOL, etc. It was inconvenient that they didn't interoperate. All those proprietary systems were wiped out by an open one.

Either users will pay with their money or pay with their data (getting ads), else the model is not sustainable. Since, messaging using Internet is being considered here as basic necessity, one could argue that Messaging apps turns to be non-profit organizations or government supported organizations (but anonymous).

Costs are very low. If the owners don't want to maximize profit, the service can very well run like this for a very long time.

You are only considering the cost of infrastructure. What about the basic salary for the people working on the application and monitoring infrastructure, who will pay for that, if not users and not ads and not government?

The team is amazingly technical. Nikolai Durov is a triple IMO gold medallists (and from Russia where selection is much more competitive than most other countries)


It's either app, or users are the product. When I was in Hong Kong recently, listening to talks about WhatsApp & Facebook, each person using these services was worth +- $48. Sub. fee? Anything like that.

Telegram is a great messaging apps. Private, desktop and web support, fast, better UI compared to WhatsApp. The problem is my friend don't know it so I have to be stuck with WhatsApp for now.

The fast follower problem.

>Secure messaging should be free for everyone. Displaying ads alongside your private communication seems out of place, even immoral

I completely agree. In this ideal situation, who pays for it? The article seems to have left this small detail out. How do you plan on getting this service to work without someone footing the bill?

It wasn't left out. You must have missed it.

Will he continue to bankroll the business -- or does he see revenue opportunities? "We will become financially sustainable at some point," he says. "It will most likely involve third-party paid apps built on the Telegram platform."

The marginal cost of messaging is low enough that phone companies are happy to offer you an incredibly low cost per formerly-private communication (landline or mobile).

while the marginal cost of e-mail is so low that we get tonnes of spam.

Who needs to pay anything more? We already pay for the infrastructure.

This question is answered rather quickly when you ask two questions:

1. What exactly are people paying for? 2. Is that actually necessary for messaging?

If the answer to 1 is servers, the answer to 2 is no: distributed or federated messaging is entirely reasonable and possible to do. If the answer to 1 is paid developers, the answer to 2 is no: there are plenty of developers willing to work on open source. Try it.

Messaging is important because it provides a person-to-person communication platform for arbitrary data. It blends in the rich interaction capabilities offered by internet with more personalized and free-form interaction offered by telephones or personal visits. Sending text messages to each other is only the first step. One can perform any kind of communication on top of messaging with proper apps. You can charge for these value added services and generate reasonable revenue.

Registration is open for Startup School 2019. Classes start July 22nd.

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