Can Amazon keep investing in AWS as losses mount?

On Thursday, Amazon Inc. posted a “gigantic” $544 million operating loss for its third quarter ending September 30. And although CEO Jeff Bezos has long shown a willingness to sacrifice profit for growth, you have to wonder how long shareholders will go along with that strategy and what impact continuing losses will have on the company’s hugely successful Amazon Web Services effort.

As Barron’s (registration required) put it:

“A two-year-old start-up can get a pass from shareholders for reporting quarterly losses, provided that its growth prospects are bright. But ‘just wait’ is a tougher sell when the startup recently turned 20, as did this year.”

On Friday, investors voted with their feet, sending Amazon’s per-share price skidding 8.3 percent to $287.06 from the previous day’s close of $313.18.

Complicating matters for AWS is the growing perception that [company]Microsoft[/company] Azure and [company]Google[/company] Cloud Platform are viable competitors to AWS backed by companies with very, very deep pockets.

AMZN Chart

AMZN data by YCharts

One long-time AWS partner thinks [company]Amazon[/company] corporate’s fiscal situation will inevitably affect non-retail efforts, including mobile devices, and perhaps AWS itself.

“This going to drag Amazon’s efforts in the cloud,” he said via email. “My feeling is that they will start by ejecting non-core, high cost initiatives like the phone, but eventually it may [affect] their ability to invest in cloud infrastructure at the levels that Google, Microsoft or even someone like IBM is prepared to invest.  This business is all about scale.” And scale costs money.

I’m sure AWS execs would disagree, but you do have to wonder how long the company can keep pouring resources into these efforts if its macro position does not improve quickly. Last year, Amazon SVP Andy Jassy told AWS Re:Invent attendees that Bezos believed that AWS could become Amazon’s biggest business. It will be interesting to see if he updates that sentiment at this year’s event next month.

Former Microsoft CEO Steve Ballmer — who clearly has a bias in this arena — still makes some sense when he said on “Charlie Rose” this week that businesses, at some point, need to make a profit.

“I like Amazon, nice company but they make no money … in my world you’re not a real business if you make no money. I get it if a company doesn’t make money for two years but Amazon is, what, 21 years old and doesn’t make money … I don’t know when they’ll make money and if I’m a share owner … It’s an expensive company, the market cap is about $150 billion, almost. Eventually, if you’re a $150 billion company, at some point you need to make $15 billion pre-tax and they’re at zero and there’s a big difference between zero and 15.”


Given all that, remember that AWS remained a bright spot in overall Amazon results, with sales rebounding and the company claiming a healthy 90 percent growth in usage.

And at a macro level, remember the landscape is littered with those who sold Bezos (and Amazon) short. Doubters should know that the Harvard Business Review just named Bezos the best CEO in the world.

So there.

Structure Show

[soundcloud url=”″ params=”color=ff5500&auto_play=false&hide_related=false&show_comments=true&show_user=true&show_reposts=false” width=”100%” height=”166″ iframe=”true” /]

The Amazon, Google and Microsoft clouds all stand to gain from the burgeoning interest in the internet of things — the massive implementation of sensors and devices worldwide that enable a broad array of applications from medicine to automotive design, in the latest Structure Show podcast.


Hosts: Barbara Darrow and Derrick Harris

Download This Episode

Subscribe in iTunes

The Structure Show RSS Feed


Microsoft’s big freaking servers make cloud wars even more interesting

Former MongoDB CEO talks about stepping down to be a better dad — and NoSQL of course

While HP splits up, married entrepreneurs build something together

Ericsson mind melds with Apcera; Talko takes on communications; and what’s up with Watson?

Launching open-source projects is easy; keeping them running well not so much