When you need to start making money, you suddenly need to a lot more engineers. Now you need developers to make sales tools, compliance, moderation, finance, A/B test engineers for growth ideas, niche features for big customers, engineers to bring down cloud costs, etc.
It doesn't make their engineering less impressive but I think we need to understand the full picture.
At the end of the day, WhatsApp needed to make money to keep their service going.
you posted the same comment two times already.
same response
They only spent $9.9 million over $10.2 million in revenues for operations.
WhatsApp’s goal is still growth, rather than monetization. Mark Zuckerberg and WhatsApp CEO Jan Koum said when the acquisition was made in February that ads aren’t the right way to earn money on messaging
They were investing on growth, an investment that lead to a $16 billion acquisition from Facebook.
I would love to replicate their same "unsustainable business"
Are you really that upset about GAAP accounting and how non-cash, stock-based compensation can lead to paper "losses"?
My point remains. They created impressive technology. But 900m users and 50 engineers needed more context. It's not like they shipped a 900 million iPhone-like product with 50 engineers. They made $200k/engineer which is pretty bad compared to more successful companies. When they needed to make more money, they would have needed a lot more engineers.
They never spent investor money from their Series A-on. After Sequoia completed the Series B, Jan and Brian forwarded them a copy of the bank account statement before the Series B financing, showing the original invested funds untouched. They were acquired less than a year later.
Whatsapp was a money-printer! It deserves that reputation.
* ability to isolate a fairly universal need that can be solved in full by software. as soon as you have to customize to meet cultural/political fragmentation you start losing orders of mangitude. once there is a need for a "human in the loop" to provide person-to-person support the "magic" is pretty much gone
* ability to deploy across vast numbers of devices. this universal canvas is generally not the case. exceptions being e.g. web standards or having an oligopoly in mobile devices
still, despite all those caveats, software is a completely different game at its heart. in principle it scales to infinity and it becomes a whack-a-mole game to figure out which factor will land things back to reality
Few hundred or thousands of users can run happily off one server. Then you can just... buy a bigger box for long time, maybe not need more for hundreds of thousands of users.
And once you put all of the effort to get highly available loadbalancing, a bunch of app servers, maybe use some distributed database or add a lot of caching for your current one you can, again, just add more and more nodes for a long time till you get to "next step".
Of course, make your app in Rails with people that have no clue about architecture and some N+1 queries and you might need to get to the scaling steps much earlier
> Imagine 50 "bricks and mortar" engineers trying to provide anything to 900M users.
They are not providing any support to the users tho. Hell, I'm pretty sure some of the best selling cars and motorcycles in history(...being probably beetle and honda super cub) had less than 50 engineers
So you're a front-end guy?
1000000000/100 = 10^7
This is a feature of software as an economic good which, particularly when further leveraged in open source ecosystems, makes it fundamentally different that any other example. Imagine 50 "bricks and mortar" engineers trying to provide anything to 900M users.