Why cloud migration is worth it? What to consider when optimizing your company's budget? What costs can be cut without compromising system performance? Read about five ways to save money with a cloud in the article by Maxim Berezin, Development Director at CROC Cloud Services.
Store is sore: Who needs clouds?
Among all CROC and CROC Cloud Services customers, there is maybe the only group not seeking to reduce capital costs – enterprises that are preparing for IPO. They need a documentary evidence of their "fat" capital assets, with lion's share being often represented by data centers. Other customers want to drop the dead weight of “heavy” IT systems either completely or by leaving only a part of them to support critical business functions. Instead, such companies resort to outsourcing and, in particular, clouds, thus killing several birds with one stone: they gain access to a powerful infrastructure, eradicate the need to update their hardware (which otherwise should be done at least once every five years), and simplify IT budgeting (this is important to financial directors). And, of course, they reduce overall infrastructure costs. According to our customers, savings may reach 30-40%, mainly when it comes to hosting services with dynamically changing load in a cloud.
Other saving options
Cloud migration helps reduce hardware dependency and local footprint. It also helps cut costs of hardware administration and other related processes. Large companies can – but often do not want – to invest in own IT staff. That is why clouds are so popular among CIOs. They simply transfer routine hardware administration functions to a provider, which, in turn, provides skilled personnel helping a customer to forget about staff turnover, recruitment, training, and retention.
No damage to reputation
In addition to direct costs of hardware and its upgrade, never forget about associated costs, such as utility and rental payments and expenses connected with searching a proper site for server capacity. According to some of our customers, this is not easy, particularly, in Moscow due to the lack of inexpensive commercial premises. In addition, available sites often do not meet all requirements for computing hardware installation. For example, in a building where a data center is to be deployed, there may be links from only one or two key ISPs and no way to add other ones. Or infrastructure and utility/communication lines in the building and around it may be so outdated that building server rooms in such conditions may result in downtime, reputation damage and, ultimately, financial losses. Let me remind you of a manufacturing company in the Moscow Region. Its office was located at the intersection of trunk viaducts, which the regional authorities decided to reconstruct. Continuous construction activities nearby caused power failures and connection losses. After few days of production system downtime, the company decided to migrate to a cloud with, among other things, fourteen (!) telecom providers available. This ensured uninterrupted access to business-critical apps and eliminated reputation risks.
Dynamically changing loads and money-saving
Let us imagine that a company has decided to go cloud. Can we learn beforehand how to reduce an average service bill? Yes, it's quite possible, and we are not talking about just chaffering between a provider and a customer. In our estimates, customers migrating their systems with variable (dynamically changing) load to clouds will harvest the greatest economic benefits. This can be virtually any service end customers use during specific periods or websites of online stores, which regularly launch promo campaigns and face visitor influx. However, if the load is generally stable, using hardware in a commercial data center and connecting cloud resources during service utilization growth spurts (online sales, Black Friday, etc.) is much more worth the money. In such a way, companies save money on IT outsourcing and insure themselves against the risk of service unavailability at peak times.
Paying for the consumed resources only
Is it possible not to pay for a cloud at all? Yes, sure! This is the very essence of cloud services. Thanks to billing mechanisms, only actually consumed resources are paid for, which is good for customers in many cases. For example, when signing a perpetual contract with a provider, a customer service developer knows that it only will run VMs when load testing is required. This means that VMs are not used in other periods and cloud service bill is zero. One more example: a company can use a cloud as a backup platform for quick switchover when its primary site is unavailable. No emergency – No payment! By the way, we strongly recommend this data backup approach to our customers since it is both reliable and cost-effective.