Are EA projects worth the price?

What are values of an Enterprise Architecture (EA) project? It might be a difficult to explain the value of EA in a word or two. However, responsibility for explaining its value lies on enterprise architects, and Anne Lapkin, Research VP of Gartner, says they haven't fulfilled their responsibility.

"Most enterprise architecture teams do really bad jobs of explaining value of enterprise architecture to the business. " (Anne Lapkin)

She points out that following points are required for enterprise architects to communicate the value of EA.

1. To understand your value prepositioning in terms of what you are delivering to your business.
2. To set up a formal communications discipline to be able to communicate that value to the business and to your other stakeholders.
3. To put in place some concrete measurements so that you can actually demonstrate that you are delivering value to your business.

So, how can enterprise architects explain the value of EA? One of possible explain is as follows.

One of well known reason of IT Project failure is lack of communication between business people and IT people. Enterprise architecuture (EA) solve this problem analyzing an enterprise from the coordinated views of the entire enterprise. Also, EA help to improve project-planning, dicision-maiking, risk-management, and etc. Well, is it too abstract? Say it is true, how much it is worth? Even only IT systems cost too much for many companies; do EA projects deserve additional investment?

It depends. EA gives us comprehensive view of current and future enterprise state from a view point of unique combination of strategy, business, and technology. On the other hand, creating an EA can consumes a lot of time and money. Thus, if the benefits from creating EA will not outweigh the cost, a company should not invest on the project. However, in some circumstances, creating an EA project will provide long term benefit to the company. Let's see some examples.

InfoWorld's Enterprise Architecture Awards is a showcase of successful EA projects.

• The 2011 Enterprise Architecture Awards
• 6 projects are awarded: American Express, Bayer Healthcare, First Data, Singapore Ministry of Education, Procter & Gamble, USAA
• The 2010 Enterprise Architecture Awards
• 5 projects are awarded: Aetna, Barclay's Bank, Discover Financial Services, Skandia UK & International, Wells Fargo

In the case of Singapore Ministry of Education, they achieved to reduce a lot of duplicative functions using SOA approach. During the project, the EA Committee reviewed every business cases and architecture. As a result of the IT system renovation, "the organization could respond with agility to the evolving education land scape and its changing policies" and "The Ministry has reduced the number of systems in needs to maintain and support by 44 percent, saving over $25 million through economies of scale".

These cases show that EA actually provide values for some projects, and its impact could be huge. Of course, these projects are the most successful cases and the value of EA varies by the project. The important thing is that enterprise architects need to understand the value of EA for each project from the aspect of business (not only in terms of IT), and tell their clients its value. These best practices above would help to identify the value of EA.

Why so many IT Projects fail?

In January 2012, Japan's Patent Office(JPO) announced that a project which aimed at renovating patent management system resulted in failure.

"Asahi reported that Japan’s Patent Office(JPO) decided to terminate their patent management system renewal project which began in 2006. They already spent 5.5 billion yen(US$70 million) and specially formed inspection committee in the organization recommended to stop the project." (asajin.com)

They must have enough resources to develop software. In fact, they spent enough time(already 6 years), money(5.5 billion yen), and human resources(best software and consulting companies).

Why on earth even such a resourceful project can fail?

Mr. Hagimoto, who worked for the government for a certain period of time in the project, analyze the cause of failure as follows.

1. The users rely on a leap of faith in IT without specifing how to use the system
2. The users did not fully analyze their workflow
3. Inadequate project management both in the vendors and the users
4. The vendor took the main role for analyzing the user's business process

He also points out business custom in Japanese IT industry, which is the relationship between system integratior(SIer) and a user company. (Publickey [Japanese only])

However, failure in IT project is not unique to Japan. In fact, a lot of IT projects result in failure every year: 10 Biggest ERP Software Failures of 2011.

Robert Frese summarize several kind of reports about success and failure in IT project. Surprisingly, according to a report, 70% of projects were not successful.

"The disturbing conclusion from this Standish report is that only 16.2% of projects were successful by all measures, and that of the 70% of projects that were not successful, Over 52 percent were partial failures and 31% were complete failures." (PROJECT SUCCESS AND FAILURE: WHAT IS SUCCESS, WHAT IS FAILURE, AND HOW CAN YOU IMPROVE YOUR ODDS FOR SUCCESS?) 

So, is an IT project naturally fated to result in fail? How can we increase the success rate? Based on Mr. Hagimoto's review, at least we have the space to improve following points.

• Defining project goals and objectives
• Stakeholders' expectation management
• Business analysis
• Project management

You may notice that all of these factors are not problems of software. These are basically organizational or human problems.  I don't know whether EA was executed in this project, but if EA wolud have been used adequately, the success rate might have increased to a certain degree. This is because EA deal with whole range of issues in an enterprise, and also it emphasis on human aspects of an IT project.

"Enterprise architecture is as much about people and social interaction as it is about processes and resource utilization. Understanding each of these aspects of an enterprise is essential to the development of accurate views of the current architecture and relevant, meaningful views of the future architecture."

Scott A. Bernard: "An Introduction to Enterprise Architecture"  

Give it some thought with EA3 Cube Framework. Patent management is one of the enterprise's business segment. "Defining project goals and objectives" and "Stakeholders' expectation management" seem to be mapped "Goals & Initiatives" layer. Business analysis would be mapped over multilayer, which might be mainly related to "Products & Services", "Products & Services", and also "Systems & Applications".  "Project management" is eliminated because it is not about an enterprise itself and has relationship with whole process of the project.
(* This is just a personal viewpoint and may not be a correct answer.)

Each vertical layer is composed "components" and each components contains substantive "artifacts", which are usually documents of analysis, requirement, design and so on. So, if they had broken down the each component, they would have crystallized some solution in the shape of artifacts.

 "Defining project goals and objectives" also defines business and organizational model after the new IT system is implemented. In other words, it defines the "future architecture" of the enterprise. Since goals and objectives are the most fundamental parts of the IT project, it will be difficult to success without clarifying them.

I can't provide detailed analysis here without any precise information about the project, and I know that the reality is not so simple. However, seeing the reasons of failure in the JPO's project, still I can't help but think that EA might have helped the project to be successful, especially in early stage of the project.

Introduction to Enterprise Architecture

"the EA is holistic and serves as an umbrella or "meta-context" for all other management and technology best practices."

Scott A. Bernard: "An Introduction to Enterprise Architecture"

When I hear about Enterprise Architecture for the first time, I was surprised that it covers broad areas of theory including strategy, business and technology. Also, application range of EA cover a lot of ground. It can be applied as documentation methodology, management program, standardized policy, and etc.

EA offers us some useful frameworks to visualize image of an organization and its information systems. The image below is one of such a tools, which is called "EA3 Cube Framework".

 

source: http://internationaleainstitute.org/ea-approaches/ea3-cube-framework/

The EA3 cube shows changiable business units. So, we may analyze a company and divide their business into functional business modules with this framework.

In a thesis, "The Trouble With Enterprise Software", service-oriented architecutre (SOA) is propoed as a solution for developing flexible and effective enterprise software. The thesis also point out that "most companies are in the early stages of a four-part transformation to SOA". EA3 will be a helpful tool to achieve this transformation.

What is interesting for me in learning EA is that I can acquire an overhead view of IT systems in an organization. I have worked for several years as ERP software developer, but I have not learned about EA. When I see this cube, I can recognize where I used to be. Perhaps, I was mainly working in "Systems & Applications" layer (also related to Data & Information and Networks & Infrastructure layers). This notion gave me a comprehensive viewpoint on an enterprise, which will enable me to design better information systems.

Another famous model in EA is the Zachman Framework. This framework is composed of matrix, which shows appropriate tools for various situations in enterprise. The rows of the matrix are layers in an enterprise ranging from executive perspective to technician perspective. The columns shows 5W1H (What, When, Where Who Why, How).

This model also gives us birds-eye view of the enterprise architecture.

Then, what is the relationship between UML, which is familiar diagram for engineers, and Zackman Framework. Is UML a part of Zackman? Well, It seems not. Zackman Framework seems only shows abstract instruction for each situations, and not mentions any particular tools such as UML. These diaglams are independently developed for different purposes.

However, some people says to understand interrelationships between diagrams will give us advantages in developing IT systems. Gundars Osvalds proposes an example of combination of Zackman Framework and UML, where UML is used to define the implementation of the architectual model in Zackman Framework. Also, Vitalie Temnenco shows examples of collaboration of Zackman, UML, RUP (the Rational Unified Framework).

These case shows some of frameworks are vary flexible, and we can choose some of them depending on the purpose. In this sense, we may say frameworks are like "vocabulary" of enterprise architects. If we learn many frameworks, we may expand range of solutions. On the other hand, if we only have poor "vocabulary", the range of solutions will be limited.

Also, enterprise architecture is an ever-improving field. Since, EA covers both of technology and business filed, it need to be updated when a new technology appears.

In 2011, Forrester announced a report named “The Top 10 Technology Trends EA Should Watch: 2012 To 2014″. The list is as follows.

1. Elastic Application Platforms Emerge.
2. Platform as a Service Crosses the Chasm.
3. Data Services, Virtualization Reach Critical Mass.
4. Holistic Integration Enables Agile Enterprises.
5. Social IT Becomes Enterprise Plumbing.
6. Improved Virtualization Sets Stage for Private Cloud.
7. Always On, Always Available Is the New Expectation.
8. Network Architecture Evolves to Meet Cloud Demands.
9. Personal Device Momentum Changes Mobile-Platform Strategy.
10. ‘App Internet’ Ushers in the Next Generation of Computing

(Source: http://horizonwatching.typepad.com/horizonwatching/2011/12/forrester-top-10-trends-in-enterprise-architecture.html)

As the same as another IT related filed, EA need to catch up with these quickly changing trends.

Nowadays, the presence of enterprise architecture has became larger. Enterprise architecture is widely adopted by major companies. Basant Mehta, Strategist of Hewlett Packard, says that any of the fortune 500 companies has separated enterprise architecture group.

(Source: YouTube)

The evolution in IT and business will make EA a fundamental activity of management of an enterprise.