One of the main drivers of cloud computing is its significant cost savings. This report by McKinsey recommends that companies should understand cloud economics and cost savings right from the beginning so that they can adjust price-performance and focus on the business value of moving to the cloud.
When it comes to on-premise storage, the cost of ownership is very different from enterprise cloud storage pricing models. The good news is that all major service providers have more or less similar cloud pricing storage models. Once you get familiar with these pricing models, you can work out a more cost effective cloud pricing strategy.
One scenario that we see often is that since companies have many cloud technologies to choose from, they end up opting for different cloud providers for various use cases or different parts of their operations. This leads to complications and complex cloud pricing models that impede efficiency and flexibility. This approach often derails companies that are well on their way to optimizing their performance and resources using the power of cloud computing.
Outsourcing cloud technologies to different cloud service providers just to cut costs undercuts efficiency and performance significantly. In order to make cloud computing work for them, companies need to look beyond common myths about the cloud price-performance and understand cloud computing from the point of view of key business objectives.
Myth 1: If it is a cloud workload, it has to be virtual
There is a common myth among companies that in order to reap the cost benefits of cloud, they should only invest in a virtualized infrastructure. This is a misconception. In fact, some workloads need bare metal or physical server options. Bare metal servers also allow companies to configure storage and network resources.
Where do bare metal servers figure in your price-performance assessment? You can, in fact, include them in your assessments for many workloads. Physical servers work well for high-throughput database systems and workloads that are performance-sensitive and also need very specific security. Physical server options can ensure that these workloads perform even better and at a lower cost than with virtualized options.
Myth 2: Cloud is always cheap, so I don’t need to match my workload to the right service
If a company is assessing price-performance for many workloads, it will find that a virtualized cloud may not often yield the desired price-performance level. In other words, when it comes to cloud computing, the cost depends on how efficiently a company is using its resources. Sometimes, to achieve optimal price-performance levels, companies may need to opt for bare metal servers, dedicated servers or even a hybrid model. It is important to assess the workloads and align them to the right cloud deployment models. Companies can also run their businesses on small, dedicated servers with more productivity and cost savings. Similarly, applications that need constant changes may work better and cost less when on the cloud.
While cloud does lead to cost savings, everything depends on how much the company understands and evaluates the costs, including the hidden costs that the vendor doesn’t always advertise. These costs differ based on performance, management and security. For example, when opting for a virtualized cloud environment, the workload performance will depend a great deal on the hardware on which it is running, which can give much higher performance if it is a new generation, high-performance hardware. Many cloud service providers charge higher rates for new generation hardware.
When evaluating price performance, companies can keep in mind that cloud service providers make it possible for companies to select the right size compute unit. This size is to a large extent based on processor capacity.
Myth 3: Opting for multicloud gets me better rates from different cloud providers
One of the biggest misconceptions that companies have is that a mix and match method cuts costs significantly. Companies want to deploy the same workload with different vendors. There are many problems with this mindset. For one, the public clouds marketplace is so competitive that there is no significant difference in costing between different vendors. Secondly, this approach does add to hidden costs. Let’s take an example. If you are opting for different vendors for your workloads, you also need to pay for management tools for all these workloads. You will end up paying for monitoring and analytics services to two vendors, not to mention the setup and infrastructure costs. It helps instead to adopt a total cost of ownership approach. You may cut some costs by using services from different cloud providers, but your total cost of ownership will be higher.
Read More: Seven Top Business Benefits of Cloud Migration
If a company want to go multicloud, it has to make sure it aligns its multicloud strategies with solid business objectives. For instance, if the company has a very mission critical workload, it can deploy this workload in multiple clouds because this means that it can avoid outage that will cost significant money.
A company can also assess and analyze the reasons why it wants to opt for multicloud. A multicloud approach can add complexity to workloads; but if a company’s multicloud strategy also saves potential losses due to outage of mission-critical workloads, then it is worth the additional cost. For many companies, business continuity costs are necessary insurance costs.
Companies have a wide array of cloud technologies to choose from. When business managers focus only on costs instead of innovation initiatives, they miss key business deadlines and objectives. In order to make the most out of a cloud computing setup’s promise to be cost-effective, a company must match workloads with the right cloud service.
Myth 4: The lowest per-unit rate gives the best overall value
Another myth that prevails is that the lowest per-unit rate gives the overall value to the company. On the contrary, once a company moves chunks of its projects and workflows to the cloud, costs increase. Increased usage may drive total costs up, even if per-unit cost may be low.
A company may work out a basic per-unit rate, but the actual cost to run a workload may work out to much more, particularly for network-reliant workloads.
Also, companies will discover that different cloud service providers price their cloud services differently. Some may make their components available for companies to pick and choose as they deem fit. Some may prepackage them into different bundles. Pricing strategies also differ from vendor to vendor, with some vendors including unlimited intra-cloud data transfer at no charge while some charge a fee based on the GB of data used. It all comes down to how cloud service providers price each component.
Let’s take the example of storage. There are cloud service providers who include permanent storage no matter which cloud services they provide, physical or virtualized. Then there are those who provide only temporary storage and charge extra for permanent storage. These aspects add towards costs significantly.
Myth 5: Shifting to cloud is primarily a cost-cutting exercise
There is no doubt that long-term storage savings from lower cost cloud storage options does work to a company’s advantage but companies need to understand how they can maximize their business value instead of simply looking to drive down costs. When companies shift to cloud purely to cut down costs, they may undercut productivity without realizing it. Instead of cost cutting, cost optimization is a more productive approach.
Unlike cost cutting, which is a one-time approach that may not necessarily lead to maximizing business value and significant cost savings, cloud cost optimization is more of a long-term strategy. As part of this strategy, companies continue to optimize both current and future costs, reviewing and evaluating costs at least once every month so that the resources match the usage.
Many companies also derive enormous benefits from a hybrid cloud, which means they shift data and applications between clouds so that they cut down on costs significantly.
Myth no. 6: We should choose a cloud service provider only based on price
Different cloud service providers configure their cloud infrastructure differently, including the components, integrations, and processes they provide. Instead of only choosing a cloud service provider for their price, companies can look at price for equivalent performance. It helps to choose cloud service providers who are able to optimize your workloads. If a cloud service provider charges less but is not able to handle the workload performance that a company expects, this will drive up costs and lead to poor performance.
When companies have preconceived notions of pricing, they fail to do their research and opt for cloud service providers who cannot provide acceptable application performance and fail to deliver the optimal performance required for workloads. When a company’s initial cloud purchase criteria is not clear and concise, it can lead to higher costs than expected, not to mention companies overpaying for many services.
Myth 7. Cloud is only about pay per use
Consumption-based storage may sound more economical but it has its own challenges. There are variable, hidden costs that add up and make it cumbersome for companies that are looking for solutions for applications that have a short lifespan. For one thing, comparing pay-per-use services is difficult because the pay-per-use phenomenon is a fairly new one and there is no widely accepted set of norms or definitions for what constitute a good pay-as-you-go cloud service. Secondly, vendors who offer consumption-based storage are leasing their equipment to companies, which means that the vendors decide when they will update or replace components. Vendors may use older technologies instead of spending money on state-of-the-art technologies and this will be detrimental to companies.
Pay-as-you-go storage services can come with hidden costs. There are logistical issues in integrating a storage system with pre-existing applications, adding to the total cost of ownership, besides the service fees. Long-term costs can add up and most companies cannot predict these costs because they cannot assess their workloads, which are for the most part uneven.
Myth 8: The main reason to opt for cloud is to reduce IT costs
Cloud computing certainly helps companies become more agile, embrace new technologies and reduce costs. Cost reduction, however, is not the mainstay of cloud business cases. Companies often stick on to this notion and lose sight of the broader business goals that the cloud can fulfill. When aligned with business goals, cloud can drive digital transformation and give a much-needed shot in the arm to the company’s functioning IT operating model.
For example, cloud can result in faster time-to-market and improved analytics, which can contribute to significant IT cost reductions and lead to revenue growth.
A good rule of thumb is to see if with increasing costs, other metrics are positively impacted too – like increase in customers. Companies should understand their cloud costs in the context of their key business objectives.
Myth 9: Moving to the cloud automatically saves money
While moving to the cloud can reduce costs in the long run and optimize workloads, it is like moving from a house with a fixed-rate water billing system to one with metered water supply. As with any pay-as-you-go model, it is all about understanding and monitoring consumption. When consumption is monitored and tracked, great costs benefits can be achieved. Companies can build a strong cloud migration plan, which will lead to more upgrades and applications that truly bring value to the business.
When companies are working out cloud storage price, they need to arrive at a cost/performance ratio by calculating performance levels and the price they pay for a workload. Price performance can also help companies determine how much it costs to deliver specific units of work. This can help them select a vendor who gives the highest return on investment.
Companies also need to use an average utilization value so that they can compare their unit costs with highest value from deliverables. They should also establish high visibility into total costs.
Summary
According to this Forbes article, executives estimate that one-third of all cloud spending is wasted. On the upside, many companies are finding ways to optimize costs, including shutting down workloads after hours. Companies are also finding ways to right size so that they can control cloud costs by turning off idle instances. They use cloud technology effectively and enjoy its many benefits. The first step is to start by gaining visibility into the cloud’s specific costs.
