How AWS is tapping the groundswell in cloud adoption

Amazon Web Services (AWS) remains the 800lb gorilla in the cloud computing market which it pioneered in as early as 2006 when it started selling IT infrastructure services to businesses in the form of web services.

It continues to enjoy that first-mover advantage to this day, even with Microsoft and Google nipping at its heels. In the Asia-Pacific (APAC) region, where overall cloud adoption is growing, AWS remains the frontrunner in four of the five APAC sub-regions, putting it well ahead of its regional rivals, according to market data from Synergy Research Group.

In an interview with Computer Weekly, Phil Davis, managing director of AWS in Asia-Pacific and Japan (APJ), offers insights on the company’s growth strategy in the region amid the groundswell of cloud adoption, how it is retaining customers and supporting them in their cloud journey.

Could you tell us more about how AWS’s business in the region has been over the past two years?

Davis: We saw strong growth globally as we went into the second half of 2020. We were excited about that growth, but we saw a little bit of a slowdown in 2021. And now, in 2022, we’re growing faster than we were before the pandemic. Our business in APJ has been following the same trend.

I think what happened was many people – with the uncertainty around Covid-19 – were trying to constrain their near-term spend. A national railway, for example, saw their ridership fall by 90% during the pandemic, but the cost to run their IT stayed the same. So, in the medium to long term, we see companies accelerating their move to cloud to get more flexibility and, ultimately, more business agility. That’s kind of how the past couple of years have been for us.

Where do you see growth coming from in Asia-Pacific?

Davis: It’s still early days in the cloud – many workloads still reside on-premise and there’s a lot of opportunity to help customers go through that cloud journey. It’s hard to pin down a growth area. We have five areas in APJ – Australia, New Zealand, Japan, Korea, ASEAN and India. In the past two years, each of the five areas has seen the strongest growth in one quarter or another.

In terms of segments, we have our SMB [small and medium-sized business] segment, including digital natives, our ISVs [independent software vendors], and then we have enterprise. We’re seeing pretty balanced growth across those segments in the region, with the SMB segment accelerating over the past three quarters or so.

In the enterprise segment, we’re seeing financial services accelerate the use of cloud as more regulators get comfortable with cloud. Manufacturing also continues to be a growth area, and that’s around everything from smart factories to applying AI [artificial intelligence] to consumer demand.

For example, Domino’s came to us with a difficult business problem of getting you a pizza within 10 minutes. The only way they can do that is to make the pizza before you actually order it, so they’re using our AI to better predict when you’re going to need a pizza.

“More often, organisations are running a vast majority of workloads, let's say on AWS, and they may have some workloads on a different cloud, but that’s not true multicloud. That’s being selective on where you place your workloads”

Other examples include helping farmers in India look at different variables to achieve better crop yields and helping airlines like Korean Air move 98% of their applications to AWS, which includes everything from flight operations all the way through to customer loyalty.

AWS has had a headstart over other cloud players in the market. How are you keeping that momentum going in the region?

Davis: It’s related to what I talked about in terms of the richness of capabilities. We continue to bring more investments to the APJ region. We just announced new cloud regions in Indonesia, New Zealand, Australia and Hyderabad in India. We’ve also announced Local Zones in Hanoi for customers that want data sovereignty and local processing for common and popular services – and still use a cloud region to get the full range of services.

We’re also partnering with telcos, whether it’s Singtel, Telstra or Globe, for things like Outposts to build out infrastructure for our customers. At the same time, we’re building up our teams here, particularly our solution architects and prototyping teams. We’re also continuing to bring all of the new services that we announce at re:Invent every year to the region as well.

More organisations today are using multiple cloud providers, often a key provider for a majority of their workloads and one or two others for other workloads. Over the past two years, have you seen customers increasing their investments in AWS vis-a-vis some of the other cloud providers?

Davis: We certainly see our current customers doubling down on AWS. More often, organisations are running a vast majority of workloads, let’s say on AWS, and they may have some workloads on a different cloud, but that’s not true multicloud. That’s being selective about where you place your workloads.

If you look at the depth and breadth of capabilities on AWS, and you want to move them elsewhere, then you have to dumb them down to the least common denominator and you don’t get all the advantages of all the innovation we’re bringing to market. And so, we see customers making selective decisions on which workloads to run on which clouds and keeping those workloads there.

Also, one of the biggest challenges we’re going to see in APJ is digital skills, and more specifically cloud skills. If I don’t have enough skills for one cloud, having enough skills for two or three clouds makes the problem harder, not better.

In the enterprise segment, do you come face to face with Microsoft most of the time as it tends to have a pretty established presence in many enterprises?

Davis: You’ve probably heard that we really try to focus on our customers and work back from there, but there’s no doubt we bump into those folks, particularly in the enterprise. And I will tell you – and this is not a Microsoft-specific comment – it goes back to the broader point that customers don’t want to be held hostage to a technology stack. They want choice and the flexibility to be able to run the things where they think it’s best and not be leveraged based on legacy technology stacks. I think that’s one of the big advantages that we can help customers with.

AWS has famously said that in the fullness of time, all workloads will move to the public cloud, but at the same time, there are these partnerships with the likes of VMware and Red Hat with solutions that enable organisations to run and manage their workloads on-premise as well as on multiple public clouds. What are your thoughts on that?

Davis: It’s not going to be one size fits all. Some customers want to run Kubernetes while other customers aren’t containerised and still want to run on a virtual environment. Some of those want to run OpenShift for their containers and others want to work with VMware environments. It’s important that we listen to customers and offer them alternatives that work with their existing tools.

Number two, there’s a big difference between modernising an application and writing a modern application from the ground up. Over time, you’re going to see more and more truly cloud-native containerised applications that take advantage of microservices with speed and agility. That’s a very different environment from a risk management system that is being moved to the cloud. You still get the cost benefits and agility, but it’s not the same degree of benefits as when you write things cloud native.

“We certainly acknowledge that there's huge responsibility associated with the privilege we've been given to work with these customers”

By the way, there are years, if not decades, of legacy code out there that will take a long time to modernise, and that journey takes several different paths. If you look at Kmart in Australia and New Zealand, they had a mainframe that they’re emulating in the cloud to run Cobol code, but they’ve also retired some code and rewritten other code into cloud native. This is where our partner network becomes very important, whether they are global and local systems integrators, or born in the cloud people like Cloud Comrade. Customers want help navigating their IT strategy and the best practices to adopt.

AWS has grown bigger over the years, with more organisations now running mission-critical applications on the AWS cloud. Over the past year, there have been some major outages in AWS that affected customers. How are you reassuring customers when these outages happen?

Davis: We certainly are very grateful and honoured that customers trust us so much every day. In fact, one of our newer leadership principles is “with size and scale comes great responsibility”. We certainly acknowledge that there’s huge responsibility associated with the privilege we’ve been given to work with these customers.

There are a number of different ways we are working with our customers to improve resiliency. Decisions on architecture are made upfront and we have reviews where we’ll sit down with our solution architects, and even our professional services team, to make sure that, depending on the availability and uptime requirements and recovery point objective, we’ve built in the right resiliency.

As an example, maybe an application can live with eight minutes of downtime a year, and you choose to run that in one availability zone. Or maybe it’s an application that needs to run in two or three different availability zones across three different regions. There are trade-offs and cost implications of running in multiple regions, so we try to understand what the customer is trying to accomplish and then architect something that meets their resiliency, availability and redundancy requirements.

The other thing we’ve been doing is to proactively look at the resiliency that customers have built into their application layer, not the infrastructure layer. A lot of customers make trade-offs that they’re not aware of and there can be surprises. We also have tools that will help them in terms of reporting and remediating problems. The last piece is we’re continuing to put security at the centre of everything we do. Every single customer I talk to has security as one of their top concerns today.

Some companies are looking at cloud repatriation because cloud services can get expensive over time for certain workloads. I understand AWS has been reducing prices for certain services and provides cost-management tools, but what is AWS doing to retain such customers?

Davis: Let me break this down into a couple of different categories. First, there are architectural decisions that can have a big impact on cost. Let me give you an example. Our Graviton processors have a significantly better price-performance ratio, so one of the things we work with customers on is, how do we help them optimise their workloads for Graviton and take cost out of the overall system?

We also look at where people spend their money. A lot of people spend on proprietary, older technologies that frankly are not really fit for purpose. Think about databases – if we can help a customer move to open source technologies and databases, they will be better fit for purpose.

There are also a number of ways customers can engage with us that can give them better benefits. There are reserved instances and enterprise discount programmes, for instance. For us, if a customer makes a longer-term commitment and helps us plan capacity better, then it helps us make better resource commitments.

Right now, there’s a lot of uncertainty in the world and customers tend to focus on the cost side of things, so we’re leaning in very aggressively to proactively have those kinds of architecture or software and commercial-type discussions to help customers.