At the end of April, we had an opportunity to attend a meeting with executives from Amazon Web Services (AWS), which delved into client case studies and approaches that AWS employs to entice new clients. The purpose of the following analysis is to provide a brief overview of AWS’s philosophy and practices, and it is not an evaluation of its competitiveness or an endorsement of its solution. For an objective evaluation of cloud service providers and their offering please see, Omdia Universe: Selecting a Cloud Service Provider 2021–22.
A consultative approach
One thing that stood out during our meeting with AWS was that cloud service providers (SPs) are increasingly employing a consultative approach to how they sell their services. To achieve this, they seem to have brought in new faces, put in place new client engagement programs, and have worked to change their culture internally. One central part of this effort is their migration acceleration program (MAP).
AWS updated the spending requirements for companies to participate in this program, lowering the spending threshold and eliminating the need for a contractual commitment. Clients projected to spend $100,000 (down from $500,000) can participate in MAP, receiving: (1) extra tools and support services, (2) some free access to AWS services, and (3) for every $1 the client spends, the cloud SP commits to spend $1 in enabling that client. During a time of accelerated costs for many enterprises this is an enticing offer.
AWS’s MAP has six building blocks, including establishing a migration methodology, determining the right tools and partners, estimating and sourcing investment, training, and other services. AWS indicated that some of its clients go through multiple MAPs and have even set up competitions between their departments.
AWS indicated that it has learned from the thousands of cloud migrations it has already executed and advises that companies looking to adopt cloud computing take it slow. AWS advises against a “big band” approach and instead suggests that companies start with the low hanging fruits, employing a “lift and shift” approach, followed by rearchitecting their workloads for maximum efficiency. The four points of advice AWS had for enterprises looking to adopt the cloud were:
- Ensure strong executive sponsorship
- Set bold and outcome-focused goals
- Build early momentum and gain expertise
- Take a portfolio-driven approach.
AWS outlined eight reasons it believes drive enterprises to migrate their applications to the cloud and eventually rearchitect them using cloud-native APIs:
- Digital transformation
- Agility and staff productivity
- Improved operational resilience, scalability, and security
- Cost reduction
- Going global quickly, M&A
- IoT, AI, and ML
- Data center consolidation
- Outsourcing changes, hardware, and software end-of-life.
We summarize four of the case studies they raised during our meeting below.
Running a banking service in the cloud
Klarna, a Swedish FinTech company, started using AWS seven years ago for application testing and development. It had already built a significant amount of its own servers and storage across six data centers. Klarna decided to “lift and shift” non-critical workloads to AWS, representing 10–15% of its total software environment.
Interestingly, when Klarna looked to expand its business and gain a banking license, AWS supported Klarna in addressing regulatory requirements in the cloud. In 2019, Klarna migrated its core banking applications to AWS’s Lambda Serverless, and currently processes 1 million transactions daily, relying on a micro-services architecture on containerized infrastructure.
Building a cloud-native data lake
Scout24, the leading operator of digital marketplaces for cars and real estate in Germany, wanted to build a central data lake and move away from its legacy Oracle environments. The multi-year transformation saw it move from Oracle to Aurora (Amazon’s fully managed relational database engine), retiring over 50% of legacy workloads, containerizing its database, and removing all dependence on Oracle Enterprise licenses. Scout24 has also grown its data lake from 340TB to 500TB since the start of 2019.
Cost reduction then innovation
Enel, an Italian energy company, migrated six petabytes and 10,000 servers to AWS in nine months. AWS brought in Accenture to align stakeholders and partners to make the migration happen. The enterprise reduced its provisioning times from 3–4 weeks down to two days and decreased its computational power costs by 21%. AWS has indicated that Enel has now started looking into new workloads and use cases, including AWS’s IoT software libraries.
Real-time business management
Zalando, a European online fashion retailer based in Germany, ran SAP applications in an on-premises data center and raised issues over application flexibility – namely not being able to implement new product ideas and roll out new features easily, quickly, and cost efficiently. Zalando migrated its SAP systems to AWS in 2015 and integrated them with 36 AWS technologies to drive insight and enable performance monitoring. AWS estimates that Zalando reduced its cost of insight by 30% and cut maintenance time by 30%.
It is very interesting to see AWS embrace a new approach in how it serves existing and prospective clients in the face of fierce competition in the cloud services space. Eliminating contractual commitments while providing migration incentives, specialized services and tools, proof of concept investments, advisory and technical guidance, and curated training looks like an enticing offer.
The one sticking point all public cloud service providers need to address is egress charges. We continue to hear enterprises express concern over data transfer costs, and while AWS did reduce and, in some cases, eliminate its charges (i.e., the first 100GB per month of data transferred out of AWS Regions is free), there is work to be done in enabling flexibility for cloud-using enterprises.
Vlad Galabov, Director, Cloud and Data Center Research