5 ms·

Basically they're a silicon valley company who've done everything that's exactly the opposite of standard silicon valley advise -- bootstrapped for first few ye

Basically they're a silicon valley company who've done everything that's exactly the opposite of standard silicon valley advise -- bootstrapped for first few years, complete apathy or even aversion to silicon valley marketing (if you live in the valley and have no foreign connection, you probably didn't even know what whatsapp was before 2012 or so), built for all client platforms simultaneously (including blackberry and Nokia), focus on non-US markets first, built out on leased hardware instead of public cloud, a simple pay-for-use business model selling to consumers (before facebook bought them), deep aversion to ads (http://blog.whatsapp.com/245/Why-we-dont-sell-ads? http://blog.whatsapp.com/245/Why-we-dont-sell-ads?) etc. The fact that their technology stack ended up being Erlang on FreeBSD is probably the least distinguishing feature about them.

riffraff 11y ago
> bootstrapped for first few years I recall whatsapp being free before it was for pay, how was it bootstrapped?
- ryanburk 11y ago
  they didn't take venture capital and instead self funded. even if you aren't generating revenue, you can bootstrap
  - flgb 11y ago
    And how did they pay expenses if they had no capital and no revenue?
    - nols 11y ago
      Bootstrapping isn't no capital, it's no outside investment. The two founders had worked for Yahoo for years before quitting, they used their savings and kept expenses low.
- dodecaphonic 11y ago
  I recall paying for it in iOS, and getting a "it's free for a year" message on Android before the Facebook transaction.
rgovind 11y ago
I interviewed with them recently. They claimed their success was due to picking up Erlang.
- dayjah 11y ago
  Our YC company (Heysan, W07) was very similar to WhatsApp except we decided to shim mobile web into the IM networks of that time. We were closing our doors as they were just getting going, and we really liked what we saw of them. At that time we had pretty much re-written everything to be erlang - it definitely makes writing IM software easy to reason about. Interoping with the IM networks was very graceful in erlang.
qyv 11y ago
To expand on this, I think a better question is why do other need so many more engineers for so many fewer users?
spyder 11y ago
"built for all client platforms simultaneously" And that's why it's strange that the article focuses on Erlang: Are all of the client apps written in Erlang?
- rootlocus 11y ago
  I believe that would be impossible. Also, the client isn't required to do a lot of parallel work, it's just an end point for the two way communication with the server. The servers, on the other hand, have to serve all the million users connected at any single moment. That requires a lot of parallelism, and that's the most likely place where Erlang was successful.
- chetanahuja 11y ago
  Not at all. The client apps were written in their respective framework/languages. Yes yes, I do believe there's some boosterism going on when their success is entirely attributed to Erlang/FreeBSD alone. Sure the stability and scalability of the backend may have been helped by using Erlang but the fact that they got to scale at all has to do with all these other business/cultural/timing/design factors that I mentioned above. Btw, the one factor I forgot to mention above was the call to use phone numbers as the namespace for user identity. I believe they were one of the early ones, if not the first ones, who eschewed email and/or social login and went straight to phone number. That made them a direct replacement for SMS in large parts of the world where SMS costs were meaningfully expensive to the population. One more example of focusing outside the US market in general and outside of Valley bubble in particular.
edwinnathaniel 11y ago
WA came out iPhone exclusive (then later on extended the offering to BB). iPhone was hugely popular in North America. BB, on the other hand, is hugely popular in Asia (and probably Europe...?) that consists of 3 most populous countries: China, India, Indonesia... guess who uses WA the most...? Support for other platforms came out very late...

riffraff 11y ago

> bootstrapped for first few years I recall whatsapp being free before it was for pay, how was it bootstrapped?

ryanburk 11y ago
they didn't take venture capital and instead self funded. even if you aren't generating revenue, you can bootstrap
- flgb 11y ago
  And how did they pay expenses if they had no capital and no revenue?
  - nols 11y ago
    Bootstrapping isn't no capital, it's no outside investment. The two founders had worked for Yahoo for years before quitting, they used their savings and kept expenses low.
dodecaphonic 11y ago
I recall paying for it in iOS, and getting a "it's free for a year" message on Android before the Facebook transaction.

rgovind 11y ago

I interviewed with them recently. They claimed their success was due to picking up Erlang.

dayjah 11y ago
Our YC company (Heysan, W07) was very similar to WhatsApp except we decided to shim mobile web into the IM networks of that time. We were closing our doors as they were just getting going, and we really liked what we saw of them. At that time we had pretty much re-written everything to be erlang - it definitely makes writing IM software easy to reason about. Interoping with the IM networks was very graceful in erlang.

qyv 11y ago

To expand on this, I think a better question is why do other need so many more engineers for so many fewer users?

spyder 11y ago

"built for all client platforms simultaneously" And that's why it's strange that the article focuses on Erlang: Are all of the client apps written in Erlang?

rootlocus 11y ago
I believe that would be impossible. Also, the client isn't required to do a lot of parallel work, it's just an end point for the two way communication with the server. The servers, on the other hand, have to serve all the million users connected at any single moment. That requires a lot of parallelism, and that's the most likely place where Erlang was successful.
chetanahuja 11y ago
Not at all. The client apps were written in their respective framework/languages. Yes yes, I do believe there's some boosterism going on when their success is entirely attributed to Erlang/FreeBSD alone. Sure the stability and scalability of the backend may have been helped by using Erlang but the fact that they got to scale at all has to do with all these other business/cultural/timing/design factors that I mentioned above. Btw, the one factor I forgot to mention above was the call to use phone numbers as the namespace for user identity. I believe they were one of the early ones, if not the first ones, who eschewed email and/or social login and went straight to phone number. That made them a direct replacement for SMS in large parts of the world where SMS costs were meaningfully expensive to the population. One more example of focusing outside the US market in general and outside of Valley bubble in particular.

edwinnathaniel 11y ago

WA came out iPhone exclusive (then later on extended the offering to BB). iPhone was hugely popular in North America. BB, on the other hand, is hugely popular in Asia (and probably Europe...?) that consists of 3 most populous countries: China, India, Indonesia... guess who uses WA the most...? Support for other platforms came out very late...