EA Principles – April 2013


The Analytical Principle is made possible by the enormous amount of Information and by new technology now available to handle this new information. The big challenge now is to use all this information to create new knowledge for better decisions and new innovations.Don Marchand, professor at IMD in Lausanne, has studied 50 global companies. Most of them have failed to create new knowledge; but increasingly companies are taking a broader view, changing their mind-set, and exploring the information now available. This requires a more open-ended analytical process that focuses on discovery of new knowledge. This new development process “Design for Use” (D4U) may be led by a Data Scientist Team. Data Science is the practice of “translating massive amount of information into predictive insights that lead to results.”

We have to know what we want to achieve and what we want to do. Change the focus from asking questions like “what” and “how much” to “why.” “Otherwise Big Data will remain a buzzword – full of sound and fury, signifying very little” concluded by James Harkin in Financial Times, March 2013.

How to Get Started?

A good advice is to start with customer, customer segments and your value propositions to these customers. Choose a specific product or a specific geographical area.Normally you have a lot of information about sales like who is buying a specific product at a specific outlet. You find that some products or shops sell better than others. Ask “why” and try to find out the answer. Set up a hypothesis. Find out which information is needed to prove your hypothesis and try to collect this information. Analyze the result, adjust the hypothesis and run your analytical process all over again!

Make a social media analysis and find out where you are able to find the information you may need in your analytical projects. The social media landscape you may search for is:

  • Facebook, Twitter, LinkedIn, Google+
  • Review sites like Angie´s List, Yelp, Urbanspoon, TripAdvisor, etc.
  • Blogs and new sites that include/encourage comments
  • Video and photo sharing sites like YouTube, Flickr, etc.
  • Search engines, such as Google, Bing, Yahoo and others

Be sure to identify and outline all of the goals and players involved in achieving your objectives. Build the business case and determine the solution architecture.

A good reason to start with customer is that customer data normally is in good order and customer is well defined in your MDM approach. Also Event Data such as customer order, delivery, and customer invoice is normally in good order and well described. When you go out externally and analyze what you find, you are able to relate to your existing information.

Be sure to update and detail your business information models with your new knowledge. That will help you sharing your new knowledge and reuse it in future analytical projects.

The Analytical Project Approach Where You Design-for-Use (D4U)

The Question is the Key

Focus more on “why” and less on “how many.” Move from “How many widgets did I sell last month on this location”” to “Why did we sell so many widgets at this location?” Which information do I need to find the answer?

Figure 1: Get to know your Customer well enough to find the right question to ask.

In an analytical project you explore available information. Based on your knowledge and your capabilities you create a hypothesis. You test, analyze and adjust it several times until you have reached a new useful knowledge to make better decisions and innovations. An analytical project requires you to ask the right questions to make the right decisions.

Superior Customer Segmentation

The starting point should be an explicit hypothesis about your customers´ needs and how you create more value for them. Once you have gathered the information required to test your hypothesis, the analysis will usually lead you to specific ideas for developing winning value propositions.

Superior segmentation – clustering your customers and prospects based on similar behaviors or preferences – can lead to much more effective targeting strategies. It means understanding more completely how you interact with your customers.

Learning to Walk

The first step you take to acquire, harmonize and mine new information sources may lead to exciting new insight. It will be important to be open to new approaches and to challenge sacred cows. You may learn things about your customers that cause you to question certain products, services or strategies.

Undertake a few pilots. Pick a product, a geographical area and a problem to focus on. Demonstrate that the return on your effort justifies the undertaking.

What Will it Cost?

What is the cost of making wrong decisions? What was the cost of Kodak not reacting quickly enough to the advent of digital photography? What was the cost of Facit in Sweden not reacting to the digital development? Once companies start investing in analytics, they almost never stop. Implementing Analytical Principles becomes a self-funding way for companies to improve their position in the market.

The Data Scientist TeamData Scientist is not a profession; it is a capability needed to perform analytical activities in your organization. Your team will need business, analytical, technology and virtualization knowledge and capabilities. It is unrealistic to think that one single person may have all these capabilities.

The main behavior challenge is to change the decision culture where more facts are used in the decision making. It is no use to make a lot of information for better decisions available if it is not used! In 2002 the American professor Daniel Kahneman received the Nobel Prize for his research in this area. After 10 years we are prepared to understand what he meant.

The team must also be able to reuse information and knowledge already established using the EA principles explained below.

The Knowledge-Creating Company

The book The Knowledge-Creating Company was written in 1995 by Ikujiro Nonaka and Hirotaka Takeuchi. They defined “tacit knowledge” (what is in your head) and “explicit knowledge” (what is written down). Their SECI-model tells us how we transform by socialization, externalization, combination and internalization.

Always try to transform the new tacit knowledge into explicit knowledge; they call it “externalization,” when transforming knowledge into a “significant corporate asset.” New knowledge becomes a corporate asset when the tacit knowledge is externalized into explicit information. Their thoughts were regarded as very provoking when published, but now they are globally accepted and you should regard them as a key success factor in your analytical project.

The Algorithm War

Data or Information

In Scandinavia we normally distinguish between data and information.

If you just get the temperature, it is data. To make it information you must know where it is measured, when it is measured or is it a prognosis, is it measured in Celsius or Fahrenheit. You may also want to know who made the measurement or prognosis. Just data is of no value; you are always looking for information and precise and trustworthy information.

When you have received the information, you want to find out how to use it when creating new information or knowledge.

The IKEA Algorithm

For many years IKEA has used weather information to calculate how many customers they may expect the following day to find out how much personnel they need. But, of course, there are more parameters that influence the need of personnel such as whether the customers have received their monthly salary, which day of the week it is, sales statistics, advertising. So the IKEA algorithm for the personnel resource has been developed during many years. With Big Data using social media, IKEA is able to further develop their algorithm.

The Airport Algorithm

For an airport it is very important to know the ETA (estimated time of arrival) for all flights. It has a big impact on the resources, the cost and the quality of service at the airport. We may all have arrived too early, when no gate is available or no one meets up to reload our luggage.

Normally it is the pilot of the arriving flight that gives the ETA. But now an American company has developed an algorithm that is much better than just listening to the pilots. It may take the weather, the traffic, the distance and other parameters into account. They are now able to sell this information derived from their algorithm to the airports. They are able to tell that other airports can save a lot of money using their information and giving their customers a better service.

So the algorithm we are able to develop may be very valuable to reduce cost, deliver better customer value or innovate new products. That is why the algorithm is kept very secret, almost like the Coca Cola recipe!

The Analytical Principle

The Analytical Principle Using an Analytical Project Approach


  • Use an analytical project approach and not the traditional IT-project approach.


  • Establish the capability to develop new knowledge based on available information using modern technology.
  • Use new knowledge to innovate and make better decisions based more on facts.
  • Make the new knowledge available as explicit information to promote sharing.


  • Establish a Data Scientist Team having business, analytical, technology and visualization knowledge and capabilities.
  • Use a project approach, where you develop theories, create hypothesis, identify relevant data, conduct experiments, refine hypothesis and repeat the process.
  • Develop a new, shared understanding of customer needs and behaviors.
  • Change how employees think about and use data.


  • See further reading

The Enterprise Architecture (EA) PrinciplesThe EA principles we have been able to introduce are very valuable in our analytical projects and our exploration of new information.

We have two overall EA principles:

  1. Manage Information as a resource and capture data only once.
    Information is a resource that is not consumed when used; that means there are no reasons to capture the same data more than once! It is just a waste of money and creates low data quality.
  2. Create simplicity to overcome the complexity.
    To have many IT systems handling the same data causes a lot of integration, adding complexity  that is difficult to handle. To maintain this complexity, we use an enormous amount of resources better needed for development.

EA Principles

  1. Create a Business-Driven Architecture
    An analytical project is more business oriented than an IT project so it is important to have the architecture based on the business and not on the IT solutions.
  2. Develop and manage an Overall Business Information Model (OBIM)
    The OBIM groups the information into 25 – 50 information groups distinguishing between Master Data and Event Data. For each Information Group like Customer or Purchase Order, there is a Detailed Business Information Model (DBIM). These models should be updated and further detailed when new knowledge is developed. Read more at http://www.tdan.com/view-articles/12655 where the Scandinavian Standard developed by DAMA Chapter Scandinavia is described in detail.
  3. Develop and manage an Overall Business Process Architecture
    We distinguish between the operational, infrastructure and innovation processes and group the activities in 25 – 50 processes.
  4. Achieve traceability from the overall architecture to the solution → details and vice versa
  5. Start at the top and keep the overview
  6. Support transformation, give assistance and share EA knowledge
  7. Relate EAT to the Business Model Canvas (BMC)

    This principle is very important when you want to discuss which are your Information Resources needed for a specific Product or your Value Proposition. The Innovation Canvas developed by Alexander Osterwalder is a great innovation in itself, giving structure to our innovations and adding value to our analytical projects. More at http://www.tdan.com/view-articles/16160

  8. Invest in reusable components
    Compare with the car and truck industries that build unique cars or trucks for each customer, only based on standard components. They have made product master data a reusable component.
  9. Develop and manage an Agreement between EA and the Projects regarding new IT Solutions
  10. Establish EA Governance supporting the desired transformations

Further Readings about Big Data and Analytical ProjectsThe website Information Management publishes useful articles on Big Data and Analytics like:

  • Exploring Big Data in small Steps, starting with social media analytics, by Todd Nash
  • What should you expect from Big Data in 2013, by Mike Gualtieri
  • Realities of the Enterprise Data Scientist, interview with Anand Rao

The Harvard Business Review has recently published a number of very interesting articles
about Big Data:

  • Big Data: the Management Revolution, by Andrew McAfee and Erik Brynjolfsson
  • Data Scientist: The sexiest Job of the 21st Century, by Thomas H. Davenport and D. J. Patil
  • Making Advanced Analytics work for you, by Dominic Barton and David Court
  • Why IT Fumbles Analytics, by Donald A. Marchand and Joe Peppard

Article in Financial Times

  • Cloud Atlas – the answer to many Big Data problems lie in the very question we ask, by James Harkin

New book

  • Big Data: A Revolution That Will Transform How We Live, Work and Think, by Viktor Mayer-Schönberger and Kenneth Cukier, published in March 2013

Key Findings and Key Success Factors

The analytical approach in a “build to use” project is like a chef experimenting with new recipes. He has all his ingredients in good order and now exactly their status and origin. After preparing the new meal, he asks his guests about their opinion and repeats his process all over until he and his guests are fully satisfied.

Every time you have defined a problem and find some new interesting and useful information out there, put it in your store and be sure to define it and describe it. Then you will be able to reuse it for other problems in the future. You will be able to fully reuse your information when creating new information and knowledge in your business.

Visualize your new knowledge when sharing it with decision makers and innovators in your organization. Make sure new knowledge is described explicitly and not just remaining in the head of a few individuals.

Try to keep your analytical initiative business driven and avoid finding problems to solve based on available technology or IT solutions. Instead, try to find the technology and IT solutions capable of solving your business problems.

Conclusions“The vast quantities of information accumulating in the cloud “can be cleverly reused to become a foundation of innovation and new services. The data can reveal secrets to those with humility, the willingness and the tools to listen,” said by Viktor Mayer-Schönberger, professor at Oxford University.

The information resource is on the cusp of becoming a significant corporate asset. “Most of all, we have to know what we want to achieve and what we want to do. Otherwise Big Data will remain a showy buzzword – full of sound and fury, signifying very little,” said by James Harkin, director of the trends agency Flockwatching, in Financial Times March 2013.

Your comments or questions to this article are highly appreciated, just send me an email.

Share this post

Eskil Swende

Eskil Swende

Eskil is main partner at IRMÊ, a Scandinavian consulting company focusing on Enterprise Architecture and the Innovation Process. He is also a partner at IRM UK, a strategic education company in London that provides seminars and arranges yearly conferences on EA, IA, MDM and BPM. Eskil is President of DAMA Chapter Scandinavia and has developed a global wisdom network of leading experts inside and outside DAMA, inviting them to give presentations and tutorials in Scandinavia. He can be reached at Eskil.Swende@irm.se.

scroll to top