Is putting end-users first a business philosophy or a design philosophy, or both? Saying that end-users come first is easy, but system designers and equipment suppliers find that putting end-users first can be hard to do.

More accurately, the end-user comes first philosophy proposes that the end-user’s needs and constraints must come first in data center and equipment designs. However, end-users should not determine what their own needs and constraints by themselves. They may not even fully recognize their own needs. In reality, the end-user is a combination of the client’s various decision makers and influencers: management, infrastructure, and IT leaders. 

Figure 1. Many participants need to help the end-user set priorities.

Several parties should help the end-users during the need-definition process. These parties include the client’s infrastructure leaders, the client’s IT leaders, design consultants, contractors, suppliers, technicians working in the space, and the client’s management (see figure 1). However, mission-critical industry professionals must strive to balance the various needs, biases, and expectations of all the participants. In that balancing act, one theme should guide the process: end-users come first.

DEFINITION PROCESS

All participants in this process (no matter the role) must subjugate their own motives to the needs of the end-user and recognize that end-user's desired functionality should drive design. It quickly becomes obvious that each customer and/or project will have specific limitations, constraints, and goals. So what are some of the functional elements that typically need to be defined in a mission-critical application? Here is a starter list:
      
  • What level of redundancy or Tier topology is desired?
  •   
  • What is the ultimate capacity of the IT and associated infrastructure of power and cooling needed?
  •   
  • Will the build-out occur all at once or can it be strategically scaled in phases?
  •   
  • What will be the virtualization strategy or will virtualization be used at all?
  •   
  • Are there any LEED certification goals or incentives?
  •   
 Every project has its own limitations and constraints, so functional needs will have to be balanced with site and project-specific constraints. Striking this balance is what makes the design and deployment of a mission-critical facility so exciting and rewarding. In this category typical constraints in-clude:  
        
  • Where is the project located geographically? Geography can play a big role in determining which power, cooling, and security technologies can be applied.   
  •     
  • What is the power capacity available at the site? Power frequently limits IT deployment, so it matters whether the facility is an existing facility infra-structure or adjoins a nuclear power plant or hydroelectric dam.   
  •     
  • How much space is available for IT and infrastructure equipment? Space affects the rack load density and whether support equipment is located on the roof or outside the building.
          
  • Are there various operational time constraints? The end-user may need this design deployed last month, not twelve months from now.
  •     
  • Lastly there is always a budget constraint. Maybe the mission is to do work within a fixed budget, perhaps the client requires certain functionality no matter the cost (Okay, I haven’t seen that project either), or maybe there is a premium for speed so the faster solution is actually the best value for the end-user.


INPUT BALANCE PARADOX

If the end-user’s needs come first, function drives design, and design must acknowledge the constraints and limitations of the project, then there must be a defined process to capture all these project elements and assess the needs against the competing constraints. Research from the classic design book, Contextual Design, by Hugh Beyer and Karen Holtzblatt, shows that nothing frustrates a client more than when they feel their needs were not heard or when a constraint was not honored. 

Citigroup’s stunning new data center in Frankfurt, Germany. The 230,000-square-foot facility’s efficient use of energy and resource-conserving design have made it the first data center in the world to achieve LEED Platinum! Photo courtesy Citigroup. Photo: Christian Richters.

Without a process or system to document these constraints, it is tempting to begin imposing ideas on the end-user without considering their needs and limitations. The design process becomes more about an engineer’s comfort level or latest design idea, manufacturer’s newest technology, or even the end-user’s unrealistic expectations. Establishing a clear system or process that defines the end-user’s needs and documents customer and project constraints creates confidence for the end-user and an opportunity for designers and suppliers to create an application specific, balanced solution.

“Experts have been telling the technology industry for years that better outcomes will result when end-users participate in the design process,” states Harold Hambrose, CEO of Electronic Ink and author of Wrench in the System, “For the most part, this has happened through standard techniques such as surveys, questionnaires, focus groups, and working sessions that ask users to describe what they do and how they believe things should work. Unfortunately, these inputs are too often taken as gospel and used to define a set of design criteria that are not the best guidelines for constructing a [balanced] system,” he added.

A community of very bright and innovative professionals, designers, and suppliers must have humility and put the needs of others first, specifically the end-user’s needs and project limitations. The temptation of pride and selfishness kick in to look at “what’s in it for me,” but in reality “what’s in it for you” are pleased customers and the reputation for balancing functional requirements with project constraints. This end-users come first principle does not mean designers and suppliers must only react to the demands of others, but rather sets the stage for them to do what they do best and gives them an opportunity to shine in the industry. Taking the lead to define and balance the end-user’s needs and constraints multiplies the design team’s effectiveness and influence. Once these targets and boundaries are established, then there is great freedom for innovation and creative solutions within those constraints. It is a design paradox-constraints bring freedom.

Yahoo's newest data center, a 155,000-square-foot facility in Lockport, New York, that uses 40% less energy and 95% less water than conventional facilities. The air-cooled data center, which can accommodate 50,000 servers, is based on an unlikely design: the chicken coop. Photo courtesy Yahoo!

FREEDOM TO INNOVATE

It is counterintuitive to think that defining constraints and limitations brings more freedom, but there is a classic psychology case study that illustrates this phenomenon. Imagine an elementary school playground with children standing near the center of the playground. In the distance are a baseball field, an empty jungle gym, and bike racks (this is a LEED-certified school), but the children aren’t running all over the place.

Seems odd at first, especially considering there is no fence. However, an actual spatial and psychological study demonstrated that when a sturdy chain-link fence surrounded a schoolyard, the children moved about freely using the entire space, even climbing on the fence. When the fence was removed, they clumped together in small groups and stayed near the middle of the playground. Without the fence they seemed confused or disoriented, they did not venture further out to the distant fields and play equipment (Nelson’s Complete Book of Stories, Illustrations & Quotes, pp. 592-593). The fence in the study represented a definite boundary line, a constraint. It was the well-established constraint that helped the students feel safe and secure outdoors. Removing it was a mistake. So when end-users establish needs and limitations designers must honor them.

But aren't designers happier when they have fewer constraints? Failing to honor a constraint is like removing the fence from the playground; it leaves designers confused and hesitant to innovate. And there is a growing majority in the industry that loves to innovate.

However, the onus to provide constraints is not just on the shoulders of the end-user. In fact, in many cases the designers and vendors need to provide constraints. Some of these constraints may be the limitation of technological capabilities, and others of these constraints come from the experience of the vendor. That is to say, there is a need for designers and vendors to clearly communicate the various technology constraints to the end-users. In fact, there may be dangers and risks-in an end-users’ proposed data center design-that the manufacturer has observed over the years through working with other clients.

Opera at IntoMobile is making the planet a cleaner place to live by opening up a data center in Iceland that runs on 100 percent renewable energy. Photo courtesy Opera.

Much of the current literature and product development is innovative and is promoting innovation lessons from other industries. The data center industry has been experiencing innovation in IT structures, power configurations, cooling strategies-even building structures and site selection.

The variations range from modular power and cooling configurations, DCIM strategies, outside air economizing, evaporative technologies, to or making chicken coops or putting your white space inside a mountain or mine shaft (see photos for examples). The opportunity and freedom to innovate is not limited, but in each of these examples the guiding principle should be the end-user’s needs and constraints come first. That being said, the data center industry is evaluating and adopting a broad range of innovative design ideas and manufactured solutions. This innovation adoption process is not unique to the mission-critical industry.

INNOVATION AND DOMINANT DESIGN

Throughout history, numerous ideas and innovations introduced into a market stick, but others fade into oblivion. Consider ac and dc power, or Windows and DOS, or VHS and Beta.

Why do most letter-based input devices use a QWERTY keyboard (see figure 2)? The underlying concept is called dominant design. At this time it does not appear that there is an overarching dominant design in the mission-critical space (which is one of the reasons our industry is currently so exciting.)

Figure 2. The Dvorak keyboard was designed in the 1930s by August Dvorak, a professor of education, and his brother-in-law, William Dealy. Unlike the traditional QWERTY keyboard, the Dvorak keyboard is designed so that the middle row of keys includes the most common letters. In addition, common letter combinations are positioned in such a way that they can be typed quickly. It has been estimated that in an average eight-hour day, a typist's hands travel 16 miles on a QWERTY keyboard, but only 1 mile on a Dvorak keyboard. Yet QWERTY is embedded as the dominant design.

“Dominant design in a product class is, by definition, the one that wins the allegiance of the marketplace, the one that competitors and innovators must adhere to if they hope to command a significant market following” (Mastering the Dynamics of Innovation, Utterback, pg. 24). The dominant design usually takes the form of a new product or set of features synthesized from individual technological innovations introduced independently in prior product solutions. Many times, the dominant design defines how something is “supposed to look and operate” in the minds of the users. The emergence of a dominant design is not necessarily predetermined but is the result of the interplay between technical and market choices at any particular time. Dominant concepts are the product of numerous experiments, technical possibilities, individual choices, proprietary positions, and sheer market momentum. Currently, dominant design features are in flux for the data center market, but individual innovations are popping up on various applications vying for more market attention and end-user acceptance. Raised floors and really cold air used to be all the rage in mission-critical facility design. Now those concepts are becoming much less common. Some of the themes that may ultimately coalesce into a new dominant design are: air and water economizing, hot and cold aisle containment, the use or non-use of water in cooling, rack power densities, expandability of systems, server level power and cooling technologies, monitoring practices, etc.

One of the interesting market dynamics of the mission-critical space is that there are really multiple types of data centers. For example, the functionality requirements for a search-based company’s data center are dramatically different than those of a large financial institution. This variation in purpose of the data center makes one dominant design less likely for the entire industry; however, it is easier to imagine a dominant design taking root in the sub-categories of data centers.

Putting end-users first creates a chain reaction of success. Having a defined system to capture end-user needs and constraints help define the project. Definition of the functionality desired leads to freedom to innovate and design. Innovation leads to technical progress for the industry and movement to dominant design standards. Dominant design leads to better technology adoption and competition. In summary, when designers and suppliers focus their energies and talents to put the end-user first everyone wins and the industry gets stronger.