As soon as internet connectivity became widely available at reasonable speed, and offered low enough latency, service providers of various kinds began offering scalable, on-demand products that were delivered over those connections.
There are now thousands of SaaS applications, available from internet giants to startups, along with services from rather fewer providers of the other two key pillars of cloud computing: platform-as-a-service (PaaS) and infrastructure-as-a-service (IaaS). Here's how analyst firm Gartner characterises these foundational cloud services in comparison to more traditional methods of IT delivery:
With SaaS, service consumers control their data, but everything else in the IT stack is managed by the service provider. With PaaS, the application layer comes into play, while IaaS consumers control everything from the OS layer upwards.
The contrast with traditional on-premises (or private cloud) IT is very clear: the capital costs associated with equipping and maintaining data centres full of physical and virtual servers, networking and storage are someone else's problem -- and large service providers such as Amazon, Microsoft and Google can solve that problem far easier than the average business can. As a customer, you just pay for what you use.
The fundamental benefits of the 'as a service' model are well known, and include: a shift from capital to operational expenditure (capex to opex), often leading to lower TCO (total cost of ownership); access for businesses of all sizes to up-to-date technology, maintained by service providers that can leverage economies of scale; scalability according to business requirements; fast implementation times for new applications and business processes; freeing up staff and resources for other projects and priorities.
Of course there are potential downsides to 'as-a-service' adoption, which include: service outages; security, governance and compliance issues; inadequate performance; hidden costs (including the cost of integrating and managing multiple cloud services, and of handling potentially large amounts of data); service provider lock-in; and customer support issues.
Most of these potential problems can be minimised with good planning and a tightly-defined SLA (Service Level Agreement), but businesses will need to remain vigilant in order to minimise them -- and also realise that public cloud deployment will not be the answer for every IT workload or business process.
From SaaS to XaaS
There's no doubt that public cloud services are a booming business. Gartner's latest forecast puts the total worldwide revenue for 2017 at $260.2 billion, up from $219.6 billion in 2016, representing 18.5 percent growth.
"Final data for 2016 shows that software as a service (SaaS) revenue was far greater in 2016 than expected, reaching $48.2 billion," said Sid Nag, research director at Gartner, in a statement. "SaaS is also growing faster in 2017 than previously forecast, leading to a significant uplift in the entire public cloud revenue forecast." Gartner expects SaaS revenue to grow 21.6 percent in 2017 to reach $58.6 billion, and at the same rate in 2018 to reach $71.2bn.
"Strategic adoption of platform as a service (PaaS) offerings is also outperforming previous expectations, as enterprise-scale organizations are increasingly confident that PaaS will be their primary form of application development platform in the future," Nag continued. Gartner expects PaaS revenue to grow 26.7 percent in 2017 to reach $11.4 billion, and 24.6 percent in 2018 to reach $14.2bn.
The highest revenue growth will come from infrastructure as a service (IaaS), which Gartner projects will grow 36.6 percent in 2017 to reach $34.7 billion, and 32 percent in 2018 to reach $45.8bn.
As well as the traditional pillars of cloud computing (SaaS, PaaS and IaaS), Gartner tracks three other broad public cloud service categories: cloud business process services (BPaaS); cloud management and security services; and cloud advertising. Here's how the analyst firm's current revenue and growth projections look until 2020:
Gartner foresees continuing growth in all sectors (and particularly IaaS, PaaS and SaaS), reflecting the seemingly relentless march of public cloud adoption. But although 'everything' is increasingly available as a service, there's a long way to go. "As of 2016, approximately 17 percent of the total market revenue for infrastructure, middleware, application and business process services had shifted to cloud. Through 2021, this will increase to approximately 28 percent," said Gartner's Nag.
Notwithstanding the potential downsides to cloud adoption outlined earlier, the sheer scale of the migration task even for organisations that are minded to go 'all-in' on the cloud is huge: multiple apps and business processes must be assessed for cloud suitability, cloud providers and services compared and selected, services architected and secured, costs estimated and governance policies implemented, services provisioned and orchestrated, and deployments managed and monitored.
Another view of the state of play in cloud adoption comes from Spiceworks' recent 2018 State of IT report on budgets and tech trends. This survey was conducted in July 2017 and is based on responses from 1,003 IT professionals from North America and Europe working in organisations ranging from SMEs to enterprises. Industry sectors covered by the survey include manufacturing, healthcare, non-profits, education, government and finance.
Although hosted/cloud-based services are currently third in the budget allocation pecking order at 21 percent, behind hardware (31%) and software (26%), it's the category where most respondents report a spending increase:
(See Spiceworks' report for drill-downs on geography and company size.)
Spiceworks' respondents spend their hosted/cloud services budget in a variety of areas, headed by online backup/recovery (15%), productivity (10%), web hosting (9%) and email hosting (9%). Bringing up the rear are developer tools, communications and desktop-as-a-service (DaaS), all at 3 percent.
When it comes to cloud-based workloads, the top categories -- both now and over the next 12 months -- are communications and collaboration, backup/disaster recovery and productivity apps:
Six of the 11 workload categories have more respondents planning to deploy in the cloud over the next 12 months than are currently doing so, namely: high-performance computing, supply chain management, R&D/engineering, software development, eCommerce and productivity apps.
The top drivers for workload migration to the cloud, in Spiceworks' survey, were: providing access to data anywhere (42%); enhancing disaster recovery capabilities (38%); enabling better flexibility/scalability (37%); and reducing the support burden on IT staff. Less important drivers were reducing capital expenditure (28%), increasing storage capacity (26%) and improving data security (26%).
The final section of Spiceworks' 2018 State of IT report looks at the adoption of a range of new(ish) technologies, which includes several offerings in the XaaS arena -- IT automation, SDS/virtual SAN, IaaS, PaaS, SDN and hyperconvergence:
The preceding forecast and survey data give a flavour of the breadth of available public cloud services and current patterns of adoption. To finish, here's a (non-comprehensive) listing of IT components and business processes that are available as cloud-based services, and the key providers:
This isn't an exhaustive listing by any means. To illustrate how far the XaaS model has progressed, consider the fact that malware is now increasingly available, on the dark web, as a service to would-be cyber-criminals.
Increasingly, 'everything' in the IT sphere can be delivered as a service via the internet. There will always be pros and cons of outsourcing versus in-house deployment, but given service provider openness and buyer due diligence, organisations should be able to allocate their workloads and business processes across these locations in something approaching an optimal manner.
However, as XaaS uptake rises and the IoT (in particular) makes its presence felt, issues like internet bandwidth and latency, and data storage/retrieval times, are likely to become ever more pressing -- along with the need to integrate, manage and secure multiple cloud services.