The AWS juggernaut slowed its expansion in the second quarter as Amazon ramped investment in building out and bringing to market new cloud services, the e-commerce giant reported on Thursday.
The 37 percent year-over-year revenue growth of the world's largest public cloud lagged estimates and fell short of previous quarters, prompting a somewhat volatile decline in Amazon's share price in after-hours trading.
"We're very happy with the growth in absolute-dollar terms," Amazon CFO Brian Olsavsky told investors during a Q2 earnings call. "We're seeing pickup from customers in usage, increased pace of enterprise migration, increased adoption of our services, particularly machine learning services."
AWS customers spent US$8.38 billion on cloud services for the quarter ended 30 June. That amount fell short of the US$8.5 billion expected by a FactSet survey of investors.
The previous sequential quarter generated US$7.7 billion in revenue.
AWS operating expenses increased by almost US$2 billion from the same quarter of the previous year.
Despite the revenue and expense trajectory, AWS became almost US$500 million more profitable than it was in Q2 of 2018.
The deceleration in growth could be attributed to lapping the first half of the previous year, Olsavsky said, when many large enterprises greatly increased usage, delivering strong financial performance to AWS.
The largest factor for operating expenses climbing to US$6.2 billion was increased budgets for adding marketing and sales resources, he said.
The technical head count at AWS—primarily engineers working on machine learning and other advanced projects—grew twice as fast as the rest of the business.
Capital expenditures, which swing greatly between quarters, also contributed to squeezing margins. A year ago, Amazon didn't require much investment in its infrastructure, but spending on data centres started to ramp in Q2—a trend that will continue through the rest of the year.
"There's a very strong investment going on in AWS," Olsavsky said.
In Q2, AWS saw a notable rise in adoption of its machine learning services, especially Amazon SageMaker, Olsavsky said.
Thousands of customers were using the provider's various ML platforms for building artificial intelligence solutions, and AWS will continue to innovate on their behalf, he said, noting AWS released more than 200 new machine learning features and capabilities in 2018.
Database services are also a multi-billion-dollar business, the CFO said, propelled by Aurora, the provider's relational database.
Amazon's overall business grew by 20 percent to US$63.4 billion.
The company's stock closed on Tuesday at US$1973.82 per share. It dropped after earnings were reported to roughly US$1926, then recovered to above US$1944 at the time of this publication.