Adam Selipsky’s announcement this past week on stepping down as the CEO of AWS effective June didn’t come as a surprise to many. This is because the possibility of it being a short-term commitment in his second stint at the cloud giant was already discussed with former CEO Andy Jassy.
However, the timing of the announcement may have caught some off guard, especially considering AWS’ recent strong financial performance and ambitious expansion plans.
While Selipsky’s leadership was commended for steering AWS towards long-term customer-centric strategies, the company has faced some recent challenges. These include layoffs, criticisms about being slow to roll out generative AI services, and not being able to steer clear of competitors like Microsoft Azure, which is slowly eating into its cloud market share, while Google Cloud remains consistent.
Matt Garman Steps Up
The one stepping into the vastly important role – Matt Garman – isn’t new to AWS. The 48-year-old has been with the company for over 18 years, in different roles, joining as an intern in 2005. That was when Jassy hired Selipsky from a software company to oversee marketing, sales, and support for the new cloud computing initiative conceptualised a couple of years before that.
Garman has been positioned to lead AWS for years, given his long tenure and reputation as a protégé and part of Jassy’s inner circle.
Jassy brought Garman into his inner circle early on in the latter’s career and moved him through different areas of the company to broaden his experience.
In his early years, Garman held leadership roles within Amazon Elastic Compute Cloud (EC2), a key pillar of AWS’s $90 billion revenue business. He helped launch Amazon Elastic Block Store (EBS), a storage service.
Garman quickly developed a reputation for a strong work ethic—infamously staying awake for two days, running point on the effort to bring services back online during a major outage related to EBS in 2011.
Many employees saw Garman’s recent appointment as the head of AWS’s extensive sales and marketing team as a clear indication that the company was preparing him for the CEO role. They believed this position was just as much about marketing and sales as it was about technology.
Immediate Threats
Garman will have to use his vast experience and role in establishing the biggest cloud provider, his understanding of AWS’ data centres and cloud products, and his recent sales experience to fend off competition from Microsoft and Google.
While AWS’ first-quarter results showed an increase of 17% in revenue to $25 billion, Microsoft reported 31% revenue growth in its Azure and other cloud services.
This was despite their revenue being half as much as that of AWS’ server rentals even after including figures of Azure, SQL Server, Windows Server, GitHub, Microsoft Partner Network etc.
On the other hand, while Google Cloud’s $9.2 billion quarterly revenue was lower than AWS’ $24.2 billion and Microsoft’s $25.9 billion, it grew at a faster rate of around 26% year-over-year.
The pace of growth of these rivals, which aim to use new GenAI offerings like Gemini and Microsoft Azure OpenAI services to narrow the sales gap with AWS, is an obvious threat.
Garman faces a host of significant decisions, such as determining how AWS will keep pace with the massive data centre expansions of cloud rivals Microsoft and Google.
Garman’s best bet would be to consider an attempt to acquire or develop deeper ties with Anthropic, which could prove to be their ace against OpenAI and Google.
AWS and Google already offer Anthropic’s LLMs to their cloud customers, and AWS has committed $4 billion to the startup to spend on AWS cloud services.
However, Garman was among the leaders who played a part in declining an offer from Anthropic to establish a close partnership similar to the one between Microsoft and OpenAI in 2021.
For at least the past several months, Garman has regularly met with Jassy, Selipsky, and Amazon board directors to discuss strategies for growing AWS revenue, including how the unit will compete in the era of conversational AI.
Other Challenges
In addition to AI, Garman has another short-term challenge: increasing the number of Fortune 1,000 companies that choose AWS when they move workloads from their own data centres to the cloud.
These customers tend to buy cloud services differently from the startups and other technology companies that fueled AWS’ early lead, as it’s harder for older businesses to move to the cloud from their data centres. They have understandable concerns about hosting sensitive data in a cloud provider’s servers.
Earlier this year, Garman reorganised the sales team around specific industries, such as financial services, automotive, healthcare, and life sciences, to customise sales pitches to these firms.
The move also aimed to address deep-rooted problems related to conflicting strategies among sales leaders and customer complaints about the quality of service they experienced from sales and technical teams.
However, Garman’s reorganisation hasn’t gone as smoothly as expected. Some AWS sellers who were previously considered generalists moved into industry sales groups based on their proximity to certain customers, not their experience in those fields.
AWS is working to get these sellers up to speed in areas that include learning how industry-specific regulations impact customers’ use of the cloud.
Making AWS More Lucrative with GenAI Services
AWS primarily sells cloud computing instead of specific business applications, which has proven problematic when competing with Microsoft, which can pair cloud services with key business offerings.
This could soon change, as AWS is leaving no stone unturned in an effort to avoid conceding any territory. Just a day before the new CEO’s ascension, AWS made two personnel changes that could help push the company towards building apps.
An outgoing Selipsky, in a Monday note to staff, announced the creation of a new group called AWS Solutions, which Colleen Aubrey, a high-ranking Amazon advertising executive, would oversee.
This new group would combine a team developing applications such as call centre services and business messaging apps with other teams developing services for specific industries, including healthcare and life sciences.
In another development, Dilip Kumar, who helped develop Amazon’s Just Walk Out technology for retail stores, will now oversee Amazon Q, a tool customers can use to analyse their companies’ data by making requests conversationally rather than using complex database queries.
Q also allows customers to develop generative AI apps, such as a tool that creates onboarding plans for new hires.
However, Q has seemingly gotten off to a slow start since AWS announced it in December. But the company continues to believe in it and has said that companies such as Toyota, GoDaddy, and National Australia Bank use Q.
Jassy also recently created a team to develop large AI models that aim to compete with those from Anthropic and OpenAI, and he took numerous AWS AI developers for the new team.
Now, it is upon the incoming boss to determine how heavily AWS continues to invest in its generative AI push, including investments and acquisitions in startups, its AI server chips Trainium and Inferentia, and the pace at which it increasingly integrates generative AI services into its cloud and data centre offering.
However, it will be interesting to watch an energised and hungry company-veteran like Garman steer AWS through this AI arms race.
