The best way for companies to save money using public, private, and hybrid cloud

Delivering a cost-effective cloud strategy is important to any business migrating to the cloud. We’ve looked at how companies can save money using the range of cloud options available on the market.

What is the best way to save money when using a public cloud?

Moving to the cloud is not just a technology shift, but a culture and people shift too. A cost-conscious cloud culture is important: all teams need to manage cloud costs and cost optimisation is everyone’s responsibility.

Delivery teams should be responsible not just for their code, but also for managing its costs, both fixed and variable. They are in the best place to take on this responsibility, as they could make the best trade-offs and optimisation decisions related to costs. 

To enable this, organisations need to provide cost visibility in an almost real-time manner, so everyone can see what’s happening and catch cost overflows before they become a flood. The team can then track various aspects of cost for its service and would be able to see how it changed over time. 

Read next: Whitepaper: Cloud optimisation best practices: Finding your cloud's silver lining

Another step would be to invest in your people by training them to understand the cost of their applications and how to optimise their environments. DevOps favours T-shaped people with a broad set of skills who go deep in one area, so financial skills should naturally be part of these secondary skills. 

Organisations need to focus on measuring cloud spend against a business metric such as total revenue, subscribers, and orders completed. This is important because it allows you to add context to the spend and change the conversation from 'cloud cost' to 'cloud return on investment' (ROI). 

With the right mindset and culture, organisations can now turn to cost optimisation that enables them to save money that they can then invest in greater innovation. The core pillars of cloud cost optimisation are:

• Right sizing – match specifications to performance and capacity;

• Elasticity – match resources to needs in both time and volume dimensions;

• Pricing models – match workloads to appropriate discounted pricing; and

• Storage optimisation – choosing the right storage tiers and type for different data sets. 

Increasingly, cloud providers are adding more PaaS (Platform as a Service) and SaaS (Software as a Service) products to their offerings. This can offer better value for money over straight migration to IaaS (Infrastructure as a Service) and can help organisations keep a lean IT footprint. Serverless or containerisation computing using a pay-per-use model can offer significant cost savings for short-lived processes.

What is the best way to save money when using a private cloud?

In contrast to a public cloud offering, the private cloud has services and infrastructure maintained on a private network, and the hardware and software are dedicated solely to an organisation. Cost considerations for private cloud are similar to those of on-premise, as both require large upfront cost to implement the hardware, software and staff resources, with maintenance and growth as ongoing costs.  Therefore, cost savings on private cloud will rely on the traditional levers such as well-structured supplier contracts with clear SLAs, effective negotiations with vendors, the flexibility to switch products as well as ensuring the use of the right technologies. 

Since a private cloud requires procuring hardware and software upfront, it will be cheaper for predictable storage and workload needs, particularly at high rates of utilisation for a long period of time. Costs can spiral if organisations cannot predict their workloads and processing needs accurately. For short-term use cases, the private cloud will have a high total cost of ownership, and with unpredictable demands, the private cloud cannot offer infrastructure with high scalability without massively over-provisioning hardware and software. 

What is the best way to save money when using a hybrid cloud?

A hybrid cloud will have cost considerations from both public cloud and private cloud perspectives. In addition to that, distributing and managing workloads efficiently between the mix is key to optimise costs. For example, ingress of data onto a cloud platform is free while egress of data out of the cloud is usually charged, so collocating frequently used data within an application or service will reduce costs of data transfers. 

Another focus area is storage. Cloud object storage is cheaper and should be used for backups and archiving for infrequently accessed data. Meanwhile, cloud-hosted virtual file services and virtual disk block storage are typically more expensive than on premise solutions. 

A typical use case to save money on a hybrid cloud is to use cloud as a disaster recover site. A cloud-based disaster recovery (DR) site can be quickly set up using automated infrastructure as code, which dramatically reduces the cost of DR as it is minimal when the cloud environment is in cold standby. 

Another effective use case is to move bursting compute to the cloud for high demand one-off events or for business continuity use cases, such as seasonal events for online shopping.

What is the best way to save money when using multiple clouds?

Using multiple clouds includes multiple factors of cost considerations. Companies must take a foundational look at how they plan to utilise different cloud providers to get best value for money and simplify operations management.  

The first factor is understanding the business value of having multiple clouds. In our experience, the most common are service offering differentiators that will give your organisation a competitive edge; for resilience, mitigating the risk of service failure or to improve global/geographical connectivity. The need to reach these benefits needs to be weighed against the potential loss in aggregated buying power for a single provider. 

You would then need to consider how to operate multiple clouds. There are cloud agnostic tools on the market that allow an operator to build services in a multi-cloud environment, which can make it easier to search for skillsets and lower the cost of having multiple tool sets.

Terraform allows users to orchestrate infrastructure builds in a consistent language across multiple cloud providers. This approach needs to be balanced however, as there are some offerings in the market that may limit the use of a cloud provider's full potential.

Networking is also a key consideration for multiple clouds, especially if you expect to be able to exchange data between your different providers. The notion of a cloud exchange means that a provider will provide private network access into the clouds network on your behalf and broker the connections back into your Wide Area Network. This avoids the complexity of managing many disaggregated connections and can offer you operational cost savings. 

Another consideration for multi-cloud options includes considering the level of data egress. As you are charged for data which leaves a cloud provider, this can become costly if you are working with large data volumes that are transferred across on-premise or other cloud providers. 

In a nutshell

Data egress, looking at cloud-cost optimisation in terms of ROI and ensuring you understand your requirements are overarching themes in any cloud migration programme. Each type of cloud solution has unique costs attached, and it is important for any organisation considering such a project to evaluate their needs in terms of scalability and necessary requirements.