This article is based on analysis originally published at techgenix.com in early 2018. Statistics and predictions reflect the state of the market as understood at that time.
The Big Three and the 2018 Market Outlook
AWS, Google Cloud, and Microsoft Azure — these three providers formed the backbone of enterprise cloud computing going into 2018. Analysts at the time expected the trio to collectively hold around 76 percent of the public cloud market during 2018, rising toward 80 percent by 2020. The global public cloud market was forecast to grow at a compound annual growth rate (CAGR) of roughly 22 percent, putting its size at approximately $178 billion in 2018, up from an estimated $146 billion in 2017.
Against that backdrop, three major themes stood out: explosive growth in cloud storage capacity, continued expansion across all three primary cloud service models, and the persistent challenge of connecting cloud infrastructure to on-premises systems.
Stellar Growth in Cloud Storage
Data volumes were accelerating rapidly across every industry sector. A Cisco survey-based report from 2016 estimated that global data stored in data centers would reach 310 exabytes (EB) during 2017, with total available data center capacity projected at 600 EB. Looking ahead to 2018, Cisco’s forecast pointed to roughly 1.1 zettabytes (ZB) of available capacity — nearly double the prior year’s figure.
Major cloud providers were expected to respond by bringing additional data centers online and deploying higher-density storage hardware in existing facilities. The practical consequences for enterprise buyers were significant:
- Organisations would be able to migrate larger portions of enterprise data to cloud platforms, reducing reliance on costly on-premises storage arrays.
- Providers would increasingly compete on customised, tiered storage offerings — hot, warm, and cold tiers — at price points accessible to mid-market businesses.
- Affordable cloud storage would encourage adoption for data-intensive workloads such as HR analytics, financial modelling, and business intelligence.
Growth Across SaaS, PaaS, and IaaS
All three primary cloud service models were on strong growth trajectories entering 2018, though each at a different stage of maturity.
Software as a Service (SaaS)
SaaS remained the most mature and widely adopted model. Analysis from Bain & Co. projected SaaS growth at an 18 percent CAGR through 2020. Established platforms from Salesforce, Google Workspace (then Google Apps), and collaboration tools such as Citrix GoToMeeting were expected to maintain dominant positions while an expanding ecosystem of vertical SaaS products addressed industry-specific needs.
Platform as a Service (PaaS)
PaaS was positioned as the fastest-growing model in percentage terms. A KPMG report referenced at the time pegged PaaS adoption at around 32 percent in 2017, with a projection of 56 percent by 2020. PaaS appeals to development and DevOps teams because it abstracts infrastructure management, letting engineers focus on building, deploying, and scaling applications without procuring or configuring the underlying compute and networking layers. Enterprises that had already extracted most of the value from SaaS tools were beginning to look at PaaS to support custom application development at scale.
Infrastructure as a Service (IaaS)
IaaS was the battleground for hyperscaler competition. A Statista estimate at the time put the IaaS market size at approximately $17.5 billion by end of 2018. AWS held the largest share, but Google Compute Engine and Microsoft Azure were expected to narrow the gap, each investing heavily in global region expansion, specialised accelerator hardware, and competitive pricing. The IaaS price war that had characterised earlier years was giving way to competition on breadth of managed services and enterprise support capabilities.
Hybrid Cloud: Bridging Cloud and On-Premises Infrastructure
Despite the enthusiasm for cloud adoption, complete migration of enterprise workloads remained impractical for most organisations in 2018. Legacy applications, regulatory data-residency requirements, latency-sensitive workloads, and sunk infrastructure investment all kept significant compute on-premises. The result was growing demand for robust hybrid cloud architectures that could connect on-premises systems to public cloud environments without sacrificing security, observability, or performance.
Key engineering challenges in hybrid deployments included:
- Network connectivity: Dedicated interconnects such as AWS Direct Connect and Azure ExpressRoute were becoming standard for enterprises that needed predictable bandwidth and lower latency between data centres and cloud regions.
- Identity and access management: Federating on-premises Active Directory or LDAP with cloud-native IAM systems was a common integration pain point.
- Data synchronisation: Keeping data consistent between on-premises databases and cloud-hosted replicas or analytics platforms required careful pipeline design.
- Operational tooling: Organisations needed monitoring, logging, and deployment pipelines that worked across both environments rather than managing two separate toolchains.
Vendors including VMware, IBM, and the hyperscalers themselves were actively developing hybrid solutions — a trend that would accelerate significantly in subsequent years with offerings such as AWS Outposts and Azure Arc, though those products were not yet available in early 2018.
Key Takeaways for DevOps and Platform Engineers
Looking back, 2018 represented a pivotal year where cloud moved from “experimental” to “default” for greenfield workloads in most large organisations. For DevOps practitioners, the practical implications were:
- Demand for cloud-native skills — Kubernetes, serverless, managed databases — was rising faster than the supply of qualified engineers.
- Platform teams were increasingly responsible for building internal developer platforms on top of IaaS and PaaS primitives, not just managing servers.
- Security and compliance tooling had to evolve to cover cloud-native resource types alongside traditional infrastructure.
- Hybrid connectivity and integration work was becoming a significant engineering discipline in its own right, not an afterthought.
The market forces identified in these 2018 predictions — hyperscaler consolidation, storage commoditisation, PaaS growth, and hybrid complexity — did broadly play out as described, setting the stage for the multi-cloud and platform engineering conversations that dominated the years that followed.