Hacker News Reader

Comment by ggg9990 | Hacker News Reader

But their WhatsApp business model was not self sufficient. They were charging $1/year which is less than the operating cost of the servers and bandwidth.

View on HN

> They were charging $1/year which is less than the operating cost of the servers and bandwidth

Are you sure of this? If they have a billion users, they're earning a billion dollars a year off of this. Considering that they don't need to store messages for an extended period of time, or run analytics, do they really need all that much?

Most people weren't actually paying. Apparently, the vast majority of users were granted free lifetime memberships (see comments here [1] and article here [2]). When Facebook acquired it, it only had $10 million in annual revenue [3] against 450 million users. Clearly, most users were jumping on board during times when it was made available for free.

[1] http://www.idownloadblog.com/2013/07/16/whatsapp-goes-free/

[2] http://www.iphoneincanada.ca/recommendations/whatsapp-messen...

[3] https://www.recode.net/2014/10/28/11632404/facebook-paid-19-...

Do you need more than $10M/year for 450M users? As I've said in another comment they barely stored anything server-side and it's known they have a very optimised stack

You do if you’ve been acquired for $19 billion. The founders couldn’t have been naive enough to think that Facebook and its shareholders didn’t have a need to monetize such an enormous investment. Facebook is not a charity, it’s a business.

Facebook monetizes Whatsapp by limiting its growth. You can easily think of all the “social” functionality Whatsapp could have added in competition with FB and Messenger.

Facebook paid 19 billion to buy up competition, as WhatsApp was a threat to the Facebook ecosystem.

That is what I am seeing now here in India, folks in my circle barely post anything to facebook now, all pictures, status messages are on WhatsApp now.

If FB pays $19B for every messenger competitor that gets traction, that's a huge incentive to make more of them.

There are network effects at play; so once a messenger has locked-in most users, it's extremely hard for newcomers to gain sustainable traction.

yeah but now you have to compete with WhatsApp to get that traction, so maybe a hard slog.

And if FB shut down WhatsApp, users will be looking to go elsewhere, and probably not FB Messenger.

Let’s say you figure out how to run a company with 100M customers with only five employees.

What stops me from getting together a few friends and stealing your business model?

At some point you have to add value just to keep the wolves at bay.

...So you'd try to solve the same problem with a different sticker to score quick cash?

I'd like you to read this and contemplate.

http://www.catb.org/esr/writings/unix-koans/ten-thousand.htm...

If you didn't bother reading and contemplating, here's the gist:

There is no wolf to fear. We are all finite beings with finite time and resources. You 'stealing' his hypothetical business model doesn't really accomplish anything except tie up your resources retracing his steps. You would be the fool reimplementing the pipe.

Now if you are the type of person who reduces life down to only pursuing the almighty dollar rather than trying to solve difficult extant problems... Well... Maybe that would be worth it to you. Even then though, he has already captured his audience, and he'll likely have a reputation advantage over your copy.

Anywho. Food for thought.

> Let’s say you figure out how to run a company with 100M customers with only five employees.

That's not how it works. I'm pretty sure you and I could write a Snapchat clone in a weekend. This doesn't imply we'll ever reach 100M users.

You could write a Snapchat “clone” but not with anything close to its performance. Snapchat has to be the most performant app I have ever seen.

Is this new? I tried it out a couple years back on a decent spec phone and it was the slowest app I've ever used. I'm pretty sure Snapchat the app was known for being sort of a wreck at some point. My daughter still put up with it but it was painful to watch her use it.

I have a fairly high end Android phone (OnePlus 3) and Snapchat freezes up all the time. Video pretty much always stutters without fail. IIRC it was the same on my Nexus 5 as well.

Network effects?

Red Queen effect

Tremendous network effects and extremely high retention

Hosting is pretty far down in the list of operating costs for a business. Employee salary, benefits, marketing etc. is where most of the money goes.

WhatsApp had 55 employees and I'd say it had no marketing to speak of.

55 * $100,000 is already $5.5 million. Revenue of $10M/year seems pretty tight to me, for a 55 person company.

Not to mention the video someone posted beloew from an engineer talking about Erlang there in 2014 has some interesting numbers. Christmas/New years is peak, and on Christmas Eve they were maxing out at 146GB/s out.

If we assume they maybe averages 75% of that over 6 hours, that's 146Gb/s * (60606)s / 8bytes/byte * 0.75 * 0.05dolars/GB = $14,782.50 for a 6 hour period (estimate).

That's highest listed bandwidth tier price for Azure, I'm sure they would pay less for a number of reasons. But let's just say that $10k for a whole day may not be out of the realm of possibility. That puts bandwidth costs possibly North of $3 million a year. Even at $1 million, that's a lot of money.

Ok but a serious company wouldn't pay per byte if they were trying to save money. They could buy that much connectivity for like 50k a month. (Though they'd need to get some switches and have a network engineer.)

And if that's audio/video Vs just text, clever NAT hole punching techniques could reduce it if truly needed.

But, it seems WhatsApp was on SoftLayer? In that case their costs might have been vastly higher.

Real costs per employee are 2-3x the salary anyways.

I find that hard to swallow for a small company size, such as being discussed here. Do you have a reference or some reasoning to support this? I'm interested to know where it comes from.

Benefits, office space, equipment, T&E, employment taxes, etc. This is an old article but it suggests about 2.7x. [1] I agree that it might be less for a startup assuming a frugal startup but it's probably at least 2x salary on average.

[1] http://web.mit.edu/e-club/hadzima/how-much-does-an-employee-... (This is old but nothing really substantial has changed except.)

I'm not sure a ratio-to-salary is the best tool here. I mean, why would a developer on $150,000 have twice the healthcare cost, or office space cost, or equipment cost of one on $75,000?

It’s just a rule of thumb and some costs do roughly scale. I’m sure there are more sophisticated calculators and spreadsheets for this sort of thing.

I'm assuming you don't run a business. 2-3x sounds in the right ballpark to me.

Why is everyone making 100,000?

That is a very conservative mean all-in cost per employee for a group of professionals. Edit to add: in the US that is.

Yes. That implies the average salary is almost certainly <$50K and maybe more like $35-40K.

Because they are developers?

Developers would cost way more than that, especially if you factor in taxes.

I started paying before they were bought, I might even have paid twice and I loved it. Would happily paid for a few famuly members as well.

Also there should be a huge possibility to increase revenue by selling API access (think appointment reminders etc as well as the whole bot ecosystem).

When Telegram started they talked about price and I think I remember being unconvinced. I wanted to pay. As we (used to) say here on HN: if you're not paying you are the product.

(Of course now we have learned that some companies charge us and still sell our (meta)data and advertise to us.)

They did a great job avoiding taking too much funding though!

In 2015, when WhatsApp had ~1B active users, Facebook said that the overall revenue from it was insignificant enough they they were just making it free. I'm pretty sure it wasn't charging the vast majority of users (which is how it got popular in developing countries in the first place).

> Facebook said that the overall revenue from it was insignificant enough they they were just making it free

Have you considered that Facebook might not be telling the whole truth?

"insignificant" compared to what, how much money Facebook makes off of every extra user (who wouldn't want to pay) on their platform? Making it free can only be a profitable move because of what Facebook could do with the data, not because a few million dollars a year is truly "insignificant".

Articles I read when Facebook bought them said they were losing money. Most reports I just Googled were subscription only or plastered with ads but here's financial sheets:

https://www.sec.gov/Archives/edgar/data/1326801/000132680114...

According to this interview the addition of the $.99 charge was to slow down growth.

He wanted to be sure they could provide a fast, reliable service and was concerned they might get swamped by an influx of users.

https://www.youtube.com/watch?v=puVxnKuFarA

According to the interview:

WhatsApp began as a "status" service. Original iPhone had no background or push notification; user had to start the app to check others' status. When Apple added push notifications to iPhone, they added this to WhatsApp allowing status to be "broadcast".

Users in Europe were paying high prices for international SMS. These users saw that broadcast status notications were similar to international SMS, but free. Some of them suggested this to WhatsApp and so WhatApp added messaging.

Then the app took off.

WhatsApp used FreeBSD, Erlang and SSDs when, according to Koum, everyone in Silicon Valley was using Linux, Apache, MySQL, Perl (LAMP) and HDDs. Compared to others in Silicon Valley, WhatsApp had far fewer servers and employees per user.

> everyone in Silicon Valley was using Linux, Apache, MySQL, Perl (LAMP) and HDDs

PHP, not Perl.

at the time of accusation, WhatsApp was running on 16 servers! Erlang is idle for communication, they had like 1 million concurrent user on each server!!!

It feels like modern web design practices have grossly skewed people’s estimates of what it costs to deliver a service. Text messages are not fundamentally hardware intensive to deliver - unless you add tracking, “machine learning”, ad delivery, and mountains of JS for serving the app itself dynamically.

Or write the service in Ruby

I adore Ruby, but I think that's fair.

Twitter's failwhale-heavy Ruby experience is a good example. Ruby and especially Rails are good examples of trading hardware for programmer convenience. This is absolutely great when you're banging out an in-house app for 100 people, and there's no problem spending $1 per user-month on hardware. But Twitter's revenue is only $0.60 per user-month, and they need to spend on things besides severs. There is a reason that they needed to switch away from Ruby to a much more complicated architecture: https://blog.twitter.com/engineering/en_us/topics/infrastruc...

I also saw some people go the other direction. They had a Java app for serving a high-volume website. But the developers had a rewrite itch, and the execs were afraid they couldn't get acquired without a more hip technology stack. They even hired a fancy consulting firm to help, but when the first version was ready to go it was incredibly slow. Like two orders of magnitude slower to render a page. The rendering times were considered normal in Rails-land, but were a real problem at volume. So they spent another 6 weeks putting in a lot of caching while the ops people ordered a bunch more hardware.

Absolutely stellar talk by principal engineer Rick Reed about scaling their Erlang setup: https://www.youtube.com/watch?v=c12cYAUTXXs.

We use erlang for managing our device messaging, I've seen the board in the office crawl up to simply obscene numbers of connected devices... on one server. Multi millions.

To be fair it's tiny little messages, I'm talking one device does maybe a meg a day after communicating constantly

There's a talk one of the engineers gives[1] (that someone else posted here so I'm watching it) about their architecture that's published in March of 2014, and in it he talks about ~550 servers, which includes 250 multimedia servers and 150 chat servers.

The only places he mentions 16 is when he talks about the "multimedia database". I think there were 16 sharded database servers.

1: https://www.youtube.com/watch?v=c12cYAUTXXs

I am personally very impressed by this, and I suspect it is one of the less-mentioned reasons they were acquired by Facebook.

I remember reading this when it came out https://blog.whatsapp.com/196/1-million-is-so-2011

BTW they were running ejabberd.

I agree that what they did with server utilization was impressive. However the blast radius was pretty huge if something went down. Also wouldn’t scale to images/video very well.

16 servers for 10^9 users doing voip and uploading vids? How's that possible?

it didn't have VoIP when it was aquired, and back then sending media was way less common.

You can offload a lot of the work to client devices themselves.

what were they accused of?

I think he meant acquisition.

Can you provide a citation that pricing wasn’t sustainable?

Having been an infra engineer previously, that pricing sounds reasonable for a messaging service based on infra costs.

WhatsApp supports free encrypted video calling and picture messaging, even for groups. No way that it was covering costs.

Why can't it? End-to-end encryption means you don't need to hold on to pictures for lengthy periods of time, and video calling can be done without a centralized service.

It does now, it did not when it was aquired.

Neither of those costs anything extra to run though. Since it's end to end encrypted with no sever backups the images and video can be communicated directly between the clients, WhatsApp just plays matchmaker.

It could, but that's not how it does, they store the (encrypted) media in their servers so that both phones don't need to be online at the same time, and also to avoid having to upload multiple times when you send to a group.

That said, they shouldn't need to keep them around for that long, unlike FB and similar.

Video and voice calling are peer-to-peer. The only cost is the signalling server. I have a similar service, and it scales pretty well.

Are you sure? They were adding 1 million users/day

http://www.thefactspeak.com/whatsapp-adding-one-million-user...

Even at $1/year - their 1.5 billion active users should be an equal amount of revenue. I'd say that's enough to cover the cost of servers and bandwidth.

But that is it. If all you want is basic messaging then you are probably right.

However, if you are trying to build a messaging platform, which is what probably FB intends to do, then it requires more investment. Today, you might think that that is where the problem. All the users really want is basic messaging. However, in 3-5yrs from today, the reality may be very different (see WeChat or LINE in APAC) and WhatsApp maybe a complete misfit in that world.

All good products evolve. For something like messaging, if all you provide is a basic product, then you risk yourself being taken over by default platform apps (such as what happened with iMessage on iOS in US). You need stronger lockins and FB is well suited to provide those.

If you make messenger users pay anything for it, you’ll lose 80% of them on the spot.

Reply on news.ycombinator.com