Adam Selipsky, AWS CEO, on stage during his keynote address for Amazon's AWS re:Invent 2021.
Photo: Amazon Web Services, Inc.

It’s time for a new cloud price war

Protocol Enterprise

Hello and welcome to Protocol Enterprise! Today: AWS is now growing almost as fast as its much smaller rivals, Google’s Andrew Moore is between the AI equivalent of a rock and a hard place and where the top executives in enterprise tech landed this week.

Spin up

AWS shed a little light on how its Graviton Arm server processor is faring with customers Thursday during its earnings call. Of the top 50 users of its flagship EC2 cloud computing service, 48 have tried Graviton; AWS said nothing, however, about how many of them just kicked the tires.

At warp speed

AWS has been the undisputed leader of cloud infrastructure computing since it invented the category in 2006, but after Microsoft and Google decided to go all-in on the concept about ten years ago, they grew much faster than AWS. Until now.

AWS revenue increased by 40% to $17.8 billion during the fourth quarter of 2021, the company reported Thursday. After years of predictions that Microsoft and Google were poised to overtake AWS in the coming years given their faster pace, that’s only slightly less than the 46% fourth-quarter growth posted by Microsoft Azure and the 44% growth recorded by Google Cloud on much smaller numbers.

There’s been a steady narrative among cloud watchers over the last several years that AWS will one day find itself overtaken by Microsoft and Google unless it adds office productivity tools, business analysis suites or similar types of software used by business professionals to move beyond its core customers, developers and operations engineers.

  • Yet nothing really changed on that front in 2021: AWS remains an afterthought when it comes to the business software market, at least compared to products like Office 365, Microsoft Dynamics and Google Workspace.
  • Part of the reasoning here is that basic cloud infrastructure services have become commodities, and therefore AWS’s first-mover advantage would wear off as buyers realized they could get the same performance and reliability from the other clouds with the business software they need as well.
  • However, judging by the fourth-quarter and full-year results of all three companies, it’s not clear buyers agree.

Microsoft expects Azure growth to increase in the coming quarter,it said last week, so it’s possible this moment in time is a blip. But it signals that we’ve entered a very interesting stage in the competition to modernize the enormous number of applications and workflows that have yet to touch a cloud server.

  • If Microsoft and Google’s bundles of enterprise software are not dissuading cloud newcomers from choosing AWS, they’ll want to consider other tactics.
  • One area where AWS has clearly followed its smaller rivals is in packaging cloud services for specific vertical markets like manufacturing or finance; expect Microsoft and Google to double down on their efforts to win over entire industries.
  • Reliability has always been a top consideration when choosing a cloud vendor, but the stakes will only get higher as the product gap narrows between the Big Three.

And there’s one tried-and-true tactic that customers love and regulators might tolerate: Maybe it’s time for a good old-fashioned cloud price war.

— Tom Krazit (email | twitter)


Dataiku is the only AI platform that connects data and doers, enabling anyone to transform data into real business results — from the mundane to the moonshot. Because AI can do so much, but there's no soul in the machine, only in front of it. Without you, it's just data.

Learn more

Unfulfilled AI worries and frustrates Google Cloud’s AI chief

Google Cloud’s mild-mannered head of AI doesn’t seem like the sort of guy whose feathers get ruffled too often. But when a promising machine-learning model prototype lingers, never to be deployed, Andrew Moore, Google Cloud’s vice president and general manager for AI and Industry Solutions, can get “really worried” about it.

He might even raise his voice when you hear him talk about it.

“It’s actually really frustrating,” he told Protocol last week. “If you're a data scientist, you're still in heavy demand, you really want to do something useful with your skills,” he said. “And you've got your report in your notebook, which says, ‘yes, I can get 20% more accurate in spotting problems before they happen. And then you think you're done. But you're actually only at the start.”

Oftentimes it’s the people – the chief technology or chief information security officers – who hold back a promising AI model from coming to fruition, said Moore. “The first part of the chasm is definitely change management,” he said. “You should never try to do a big AI deployment without some very senior leader, ideally a CEO, saying they want to see it.”

But AI models morph after they’re put in production. Moore said incorporating model monitoring and explainability tools into the ML operations process as “just a general piece of the infrastructure” can help. “After you've launched your machine-learning model, it's got to keep on living. And so you do need to be monitoring it,” said Moore.

— Kate Kaye (email | twitter)

Upcoming at Protocol

The cloud let chief information officers become more strategic and focused on building business value instead of managing IT as a cost center. Then the pandemic thrust CIOs into an even more central role: They became crucial leaders when managing remote work, navigating abrupt changes in business operations and accelerating digital transformation.

So what’s next for CIOs? Join Protocol’s Tom Krazit for a virtual event on Feb. 8 at 10 a.m. PT in discussion with four amazing IT leaders: Rob Carter, CIO, FedEx; Chris Bedi, CIO, ServiceNow; Sheila Jordan, Chief Digital Technology Officer, Honeywell; and Vittorio Cretella, CIO Procter & Gamble. RSVP here.

Enterprise moves

This week, Google lost more members of its Ethical AI team while executives in sales and data left for startups. Meanwhile, Okta, Boomi and Workday all appointed new C-suite members. At companies large and small, here’s what’s happening with the people of enterprise tech.

Alex Hanna and Dylan Baker left Google’s Ethical AI team to join Timnit Gebru’s nonprofit.

Carolee Gearhart joined Gympass as CRO after previously leading sales at Google.

Debanjan Saha joined DataRobot as president and COO. Saha was previously VP and GM of Data Analytics at Google, and held leadership roles at AWS and SoftLayer.

Brett Tighe was appointed CFO at Okta. Tighe was previously Okta’s SVP of Finance and treasurer, after working in the finance organization at Salesforce for more than a decade.

David Meredith was named CEO at Boomi. Meredith joins from Everbridge, where he was CEO and board director.

Barbara Larson took the reins as Workday’s new CFOthis week. Larson formerly led finance functions at Workday, VMware and TIBCO.

Patrick Blair was appointed president of the Americas at Workday. Blair was formerly the president of CRM at C3 AI.

Ulrik Nehammer joined ServiceNow as president of EMEA. Nehammer formerly led the APAC region for Salesforce.

Keyvan Esfarjani was promoted to EVP and chief global operations officer at Intel. Esfarjani has been at Intel for over a decade and previously led memory manufacturing.

Thierry Carrez was named GM of The Open Infrastructure Foundation. Carrez previously held leadership roles at the Open Source Initiative and the OpenStack Foundation.

Mina Naguib joined Hivestack as CTO. Naguib was previously chief architect for Samsung Ads.

Around the enterprise

Google Cloud unveiled a new, free version of Workspace that will allow users to bring their own email domains to the service while still collaborating on Google Docs and other services across small teams.

The Department of Homeland Security created a new cybersecurity review board modeled after the National Transportation Safety Board’s accident review process, and it will start by looking into the Log4j vulnerability.


Dataiku is the only AI platform that connects data and doers, enabling anyone to transform data into real business results — from the mundane to the moonshot. Because AI can do so much, but there's no soul in the machine, only in front of it. Without you, it's just data.

Learn more

Thanks for reading — see you tomorrow!

Recent Issues