There is little doubt cloud offers businesses huge economies of scale that drive down the Total Cost of Ownership (TCO), a fact reflected in its pricing. Typically considered ‘cheaper’ than on-premise infrastructure, many businesses are making cloud the foundation of their IT set-up. Cloud is popular because it takes away the challenges of managing traditional costs associated with IT – people, power and hardware – and replaces these with simpler ‘utility costs’ measured against consumption.
The consumption model and the ‘on-demand’ nature of businesses’ technology are growing in popularity, offering a ‘simpler’ way of conducting business. Further, the price war raging between global cloud providers such as Amazon, Google and Microsoft offers tempting savings; indeed, research has found entry-level cloud prices have fallen by 66% in the last two years. But despite the apparent simplicity of cloud, there is no one single way to pay for it and pricing can be difficult to compare across providers and with on-premise costs. This complexity can lead to unexpectedly large bills and prevent organisations realising the cost savings they originally envisaged.
Setting your own price benchmarks
Almost all cloud computing pricing is derived from usage against three core IaaS (Infrastructure-as-a-Service) metrics – compute time, storage and network bandwidth. With compute time, pricing is generally by capacity (CPU cores, VM sizes) but may be metered by the minute, hour or day. There may also be a choice over the underlying CPU type and server architecture, which can affect performance. Equally, the type and performance of storage a business chooses is critical in ensuring its applications run effectively.
Accurately predicting these cloud costs can be an inexact science. This is because often there are no fixed limits on spend, more staff with access to self-service IT, and automation that scales resources automatically. These factors can lead to the purchase of large amounts of additional compute or storage resources, which can incur expensive charges outside of a traditional monthly accounting system. Similarly, PaaS (Platform-as-a-Service) services can be complicated to cost; pricing is based on service specific capacity metrics, per-transaction or per-function charges, rather than traditional per-user charges.
With such a range of options, businesses must create specific benchmarks for their infrastructure and for individual applications, to avoid unanticipated spend. For many businesses, this is easier said than done. This is particularly true for IaaS or PaaS, where carrying out accurate modelling is difficult without an application prototype to run tests with. In this instance, it can be useful to engage with a third-party that can create representative real-world tests to assess cloud readiness – ensuring CIOs have the knowledge to make informed decisions on cloud.
If you opt for SaaS, ensure flexibility is built-in
For the most part, Software-as-a-Service (SaaS) appears pleasantly simple, with per-user charges offering a simple way to see what a provider offers. SaaS is also fairly easy to model, and any extra costs, for example additional document storage, are usually made clear. Further, for many SaaS offerings – Office 365 is a good example – the price of the cloud service is not significantly higher than the on-premise licence. This is despite delivering an entire service against an SLA, rather than just software, so SaaS will normally be a more affordable option.
Though when it comes to SaaS, there is a trade-off to be made. Most vendors insist on a minimum 12-month term, and offer more aggressive SaaS pricing in exchange for a fixed volume with a longer term commitment. However, this significantly limits the flexibility of a SaaS model. This could cause a future issue if CIOs need to scale capacity down again, or switch to a different offering if business needs change – for example, switching between different editions of Office 365 when extra features are no longer needed, such as Business Intelligence or telephony tools. SaaS contracts are gradually changing to include flexible Pay-As-You-Go models, but businesses must ensure contracts allow for flexibility dependent on business needs – or they may find themselves paying over the odds.
Making the price war work for you
The type of Service Level Agreement (SLA), backup or Disaster Recovery (DR) on offer also varies considerably across providers. These aspects of cloud contracts are different to on-premise solutions, so it’s important CIOs do not make assumptions about what a service will deliver and how it will be supported. To do this accurately and take advantage of the price war, businesses should model cloud applications or a migration before undertaking changes, to avoid unforeseen expenditure – or worse still, a loss of data or service.
Cloud prices will continue to fall from the steep drop-off we’ve seen over the last two years; despite margins being thin, there is still intense competition to acquire and retain customers. Though the ‘big’ cloud vendors are fairly prescriptive with contracts, there is a broad cloud marketplace, and smaller providers and cloud resellers expect to negotiate.
With a clear cloud strategy in place, businesses can make the price war work for them; with certainty around plans and usage, there are significant additional discounts available for larger volume spend, term and capacity commitments. Cloud is undoubtedly the future of IT, and with careful planning and knowledge of the landscape, businesses can embrace the benefits of cloud while avoiding unexpected costs.
