Taking a Fresh Look at the Data Center Build vs. Lease Decision

  • Home
  • Taking a Fresh Look at the Data Center Build vs. Lease Decision

When it comes to expanding data center space, the “build vs. lease” analysis has been around for quite some time.  For years, business executives have struggled with whether to build their own data center or buy space from a colocation provider.  How should these decision-makers determine what’s best for their companies today?

The build vs. buy decision can be based on a number of criteria.  Many companies look solely at the numbers.  In some cases, emotional factors creep in.  But in the end, the best decision is based on two things:  which method produces the most strategic advantage for the business and which strategy creates less of a drain on company resources.

To arrive at the best decision, executives should objectively answer these questions:

  • Will building its own data center provide a competitive advantage for the company?
  • How much control does the company need to have over facility operations?
  • Does the company have the necessary expertise to build a data center facility?
  • Does the company have the time to wait for a major construction project to be completed or do they need space now?
  • Does the company have the financial resources to fund construction and operation of a data center?
  • Could funds earmarked for a data center project be used to accomplish other strategic goals instead?
  • Does the company have the in-house talent to manage a data center operation or will it need to hire additional resources?
  • If the data center is moved to a colocation provider, can internal staffing resources be re-focused on core business initiatives?

The Risks Involved in Building 

Typically, very large companies in the technology space and/or those who want to maintain full control over facility operations choose to build, own and operate an in-house data center.  But several risks accompany this strategy:

  • Site Selection –Knowing where to build a data center takes special      consideration.  A suitable location requires a thorough evaluation of many factors, such as the cost and availability of power and water, access to carriers and growth potential.Site risks need to be identified as well.  For example, is the location in an area at risk for seismic activity, tornadoes, flooding or other natural disaster?
  • Team Assembly — Data center construction requires highly specialized expertise that few companies have in-house.  Construction requires outside professionals such as architects, engineers, general contractors and subcontractors.  The building team must thoroughly understand the latest      technologies and practices involved in data center design.  And hiring a team capable of building a sophisticated facility is extremely costly.
  • Availability Expectations –Building an in-house data center requires an understanding of availability, redundancy, reliability, uptime, tier levels and fault      tolerance.  These terms are often used interchangeably and can be misunderstood.It’s important the organization builds the data center to the required level of availability. They must acknowledge the cost trade-offs they’ll need to make.  In other words, the more “available” the environment needs to be with redundant components and systems, the more it will cost.  However, company executives need to acknowledge no environment will be absolutely bullet-proof.  Even the best designs and equipment fail.
  • Operational Capabilities – In addition to the level of infrastructure redundancy designed into the facility, operational and maintenance procedures must be established.  In addition, qualified support staff should be available around-the-clock to ensure continuous availability.  The company should expect to incur substantial costs for specialized personnel  to manage mission critical systems including power and cooling.  Plus, skills shortages may exist in certain positions.
  • Energy Efficiency – A company can build a data center today using the latest sustainable building practices and most energy efficient power and cooling systems.  Yet, as new technologies surface, the facility can become less efficient and more costly to operate down the road.  A new building’s energy systems can become outdated rather quickly.
  • Scalability and Capacity Issues–Technological advancements and changes in computing equipment happen continuously. With the constant move to higher density equipment and improved  energy efficient systems, older data centers might find it difficult to keep pace.In addition, in-house data centers may run out of capacity sooner than planned.  During the design phase of the building project, a company may have made a trade-off between extra capacity for growth and the available budget.It’s always difficult to plan and pay for growth when a facility is being built.  Companies need to walk a fine-line between building a flexible data center ready to accommodate anticipated growth and initially paying for under-utilized space and systems.

The Move to Outsourcing

Not too long ago, building a company-owned data center was the default decision for many businesses. Today, companies question the need to build, own and operate their own data centers.

Many businesses find the “buy” decision offers more flexibility and cost efficiencies.  Designing, building and operating a data center is a colocation provider’s core business.  And they can almost always do it more economically than companies not in that business.

Companies choosing to go the “lease” route do so for a variety of reasons:

  • Lower Costs – When a company builds its own data center, it incurs huge capital expenditures.  On the flip side, colocation space requires a monthly operating expense.  The lower upfront costs cause many businesses to favor a colocation strategy.  When they crunch the numbers, CIOs in most scenarios find outsourcing less expensive.
  • Faster Timeline –Building a data center facility typical takes 12 to 24 months.  On the other hand, a colocation option takes only a few weeks. If capacity is urgently needed, colocation may be the only viable choice.Rather than invest the time and money into a construction project, many companies are turning to colocation providers for their data center space.  They don’t have to wait for a facility to be built, nor risk the many pitfalls that can happen during a construction project.
  • Greater Expertise –A colocation provider often knows more about building data center facilities than the average company. After all, it’s their business.Providers know how to build and operate the most environmentally efficient facility possible.  They understand cooling requirements and can use computerized modeling of fluid dynamics to determine optimal placement of infrastructure across the data center floor.And, they have experts on staff to manage every aspect of a data center’s construction and operation 24/7.  Although some companies may find it difficult to give up control of their site, they may not be equipped to maintain a facility properly from a cost, expertise and responsibility standpoint.  Colocation providers have access to large pools of technical resources to deal with emergencies and minimize downtime.

Once a business weighs all the costs, risks and benefits of both strategies, it can make a decision on which way to go.  Colocation provides a “best of both worlds” scenario for many companies.  They still own and control the hardware, yet they don’t have to worry about things like floor space, cooling, power, cabling, fire suppression and physical security.

After making a “buy” decision, companies need to thoroughly evaluate potential providers.  At a minimum, they need to investigate the providers’ facilities, procedures, operating history and financial strength.

In today’s market, colocation definitely has the wind at its back.  As companies increase their data center space in the coming years, recent studies confirm the amount they actually own will decrease.

For more information about colocation services, visit https://cyrusone.com/.