RUSH-data enabled through Enterprise Architecture

In a turbulent market an agile organization requires RUSH-data. RUSH-data stands for Realtime, Unfiltered, Shared and Holistic data. An organization with access to RUSH-data has the building blocks to become agile and act upon market changes with unprecedented speed. It does not, however, guarantee that it moves faster and more vigorously than others.

• How do you enable RUSH-data? 
• Which parameters to tune? 
• How can Enterprise Architecture help setting the level for where to extract the data and when? 

First an example to get the thoughts flowing. Zara, the Spanish fast-fashion retailer, co-locate staff from design, manufacturing, supply-chain and store operations in units of 240 people per cohort. That enables the people to share unfiltered data in realtime enabling them to get an holistic view of the current state. That enables them to respond much quicker to market demand and get new designs in stores within 3 weeks instead of 3 months for competitors. The cumulative effect over the years has led Zara to outpace their competition in growth and profitability.

Where does Enterprise Architecture come into play? Isn't this a case for not using Enterprise Architecture and trust oral word-of-mouth instead of complicated ERP systems? Well, it could be if you don't succeed. But we've done communication for thousands of years and still only Zara manages to leverage human communications. So how can Enterprise Architecture help here?

First of all you must have a culture of information sharing and thirst for more data. Otherwise no system in the world can save you. Prepare for a funeral ...

Then I believe Enterprise Architecture can help in these ways for each element of RUSH-data:

• Realtime
• EA sets the non-functional requirements for gathering data. Is it set for real-time or only providing snapshots? How does the snapshots accumulate across IT systems? Will it distort data accuracy and age?
• Which data needs to be available in real-time to enable RUSH-data?Aggregation of data across systems often leads to distortion disabling real-time attributes. When designing the EA design for RUSH-data.
• Unfiltered
• Is data available from its sources or only aggregated in reports with conclusions drawn? Can anomalies be traced or is it averaged and crunched to nonrecognition? Trends are important - however trends indicate strong movements. Developments grow from small significant anomalies that needs to be traced and identified. 
• Shared
• Collaboration and sharing of information is crucial in todays turbulent markets. A large component is to have a culture of sharing. Assuming a sharing culture, then how can EA drive sharing? Through assigning roles and ownership of data it will be much easier to decide upon how and to whom information should be available and in which format. Through an SOA information can be both unfiltered and shared in the right format to each person.
• Holistic
• EA is very good at identifying information ownership throughout an organization. Which master data combined with realtime information from the source will give a holistic view of how we're doing right now? Work on reducing the information size to craft the data-set needed to get the holistic view. Then create services that supply that information in a dynamic manner. Through the SOA the information can be shared, unfiltered access to raw data in realtime. RUSH-data made real through Enterprise Artchitecture.

Next thing to do is how to visualize RUSH-data. What's your opinion on the best ways to make visualization simple, fast and collaborative?

More reading:

• Don Sull's blog on RUSH-data: http://blogs.ft.com/donsullblog/category/rush-data/
• Matthias Scholander on the CIO Perspective of RUSH-data: http://www.matthiasscholander.com/2010/08/strategic-agility-the-cio-perspective/

Enterprise Architecture creating the engine for network-based business models

In a world where the network-based organizations are creating the new business models, Enterprise Architecture play the key role in making it all happen.
Value chains represent all activities performed to create the value experienced by a customer. Michael Porter has described this in his 5 Forces where the relative strength of each participant is reflected in the amount of profit they can extract from the value chain.
The business model is the engine in a firm's network-based strategy in how to cooperate with others to create value. The business model therefore plays a crucial role in determining how much profit a single firm can extract from the value chain.
In a value chain there are four sources of value creation; Novelty, Lock-In, Complementaries, and Efficiency.

Two questions that struck me are:

• How can a single firm control and steer its business model in a network-based value chain? 
• How can Enterprise Architecture enable the business model to become the engine through leveraging the four sources of value creation?

I think the answer to both questions can be found in using Enterprise Architecture, EA, to guide and steer successful execution enabling the four sources of value creation. Most transactions and interactions, if not most of the value, are created through IT systems. Enterprise Architecture will enable an organization to identify the most crucial parts of its IT system that creates the four sources of value creation. Once identified the organization can through Enterprise Architecture focus its effort on each of the four sources:

• Efficiency
• Integration with other partners done quickly and with no effort. Finding the best suited product with low search costs. Reliable lead-times and costs.
• EA focuses on the business processes and information exchanged. Thus it is possible to remove bottlenecks and based on which levels of efficiency creates the most value put the efforts there.
• Complementaries
• Flexibility in terms of vertical and horizontal integration. Offer additional services and products that complement the product. 
• EA will based on which types of vertical integration is available enable fast and smooth integration. Information necessary to provide data about available complementaries, especially those that change over time, is identified through EA.
• Lock-in
• Raising the switching costs through efficient loyalty activities. Recognizing me as a customer. Not just discounts but more importantly access to improved information, priority, customization, etc.
• EA identifies which information and business processes that drive the most customer loyalty. Providing that information through the different processes and through the value chain enables larger shares of profit to be extracted.
• Novelty
• Enable flexibility in terms of transaction structure, transaction content, offerings through network of partners and complementaries.
• EA does not only focus on the current state of affairs, but more importantly where the organization is heading. Areas where Novelty are important will be identified and necessary information, business process and IT system support is created.

I think you should ask yourself: What is you position within the value chain? How are we utilizing the four sources of value creation within our current business model? Which IT system support is needed?

Further reading:

• The Network Challenge: Strategy, Profit and Risk in an Interlinked World by Paul R. Kleindorfer
• TOGAF 9 A Pocket Guide by Andrew Josey

Designing experiments for succeeding or failing in a turbulent world

How to design experiments that surface flaws in your mental maps about what the future holds? Succeeding in turbulent times requires the commitment to action while knowing that your map of the future is flawed. Running experiments is the single most important activity singling out what will work and what doesn't. Based on the knowledge gained through the experiments and your increased understanding you can redesign your mental map and increase the likelihood of succeeding.

There are two domains we need to understand better to succeed. That is to identify deal killers and knowing what we are betting on.

• Identifying deal killers. Which events will kill your idea or business?
• Knowing what we are betting on. Which unique things are we dependent on to succeed? What will drive customers to our offering? Not the inverse from deal killers.

Getting the cold reality of flawed mental maps enables us to spot weaknesses and opportunities we didn't know beforehand. Experiments lets us under control and with limited bets get a better picture of how our mental map of the future actually fits with reality. Here are five ways of running experiments:

• Partial experiment. Test one attribute at a time. Test parts of a deal killer or bet.
• Design holistic experiments. Test multiple variables and the interactions between them on a small scale. Could also be a test launch before a larger roll-out.
• Explicitly consider trade-offs. Cheap Fast or Certain. But not all three at the same time. Limited budget? Go for cheap on the expense of certainty. Need it fast? Put in more money to match the speed and certainty needed.
• Stage your experiments. Start small and try out the major killers or bets early. Once decided - go all in for that variable and execute the experiment. In between experiments there are room for discussions - during execution it is focus on results. After each experiment evaluate and decide on next step based on results. Don't hesitate to change your mental map - after all that's what the experiment is all about. Validating and finding more/better facts for the mental map of the future.
• Avoid experiment creep. Stick to what you decided during an experiment. Don't let findings during the experiment let you deviate from the set experimentation plan.If the results are not as expected - don't interpret them in the most positive way. Change the mental map instead.

How are you running your experiments and what value do you get out of them to clear the fog of the future?

More to read:

• Don Sull: The Upside of Turbulence
• MIT Sloan: The 4 Ways IT is Driving Innovation. How IT helps drive experiments.
• HBR: How to design smart business experiments?
• Dan Ariely: Predictably Irrational

Has Enterprise Architecture reached its Plateau of Productivity?

Enterprise Architects pride themselves with driving IT from a business perspective. Are they doing that or is it just another case of technology driven IT? This time with structure and control as underlying forces.

What shall Enterprise Architects do to really benefit their organization and add value?
There are two main drivers that you as a business leader have to balance between; Exploitation and Innovation.

• Exploitation. Exploit any given opportunity to the maximum. Be as efficient and operate with low cost. The IT delivery itself should of course be Lean. But more importantly it should enable other departments to be lean in their processes and operations. Embrace possibilities. Don't be too reliant on standards, especially IT standards. Let each business area decide what is best for them in terms of functionality, not in which system to use. 
• Innovation. Innovate within the IT delivery. How can we help the business being more efficient? Have IT drive Innovation, from idea generation to replication of new products, services and business models across geographies. Enable exploitation of new innovations through effective and efficient IT solutions. What should they look like? Well, be close to the business and listen, invest in agile IT support systems that are evolving over time. Calculate at least 25% of IT costs to be spent on new development, not just IT operations and roll-out.

The aim for an Enterprise Architect is to enable, guide and steer the IT delivery between these two extremes. Pushing Exploitation to be as efficient and extract as much profits out of any given opportunity. Driving Innovation to create new opportunities and being agile enabling new sources of revenue.

The most important things to get right are then:

• Interfaces. Let interfaces drive the enterprise integration, not standard systems. Service Bus is a technological solution to a business problem. Get the business solution right then the technology will solve itself. Having interfaces, common data models and master data drive integration will definitely drive integration across business domains in large enterprises.
• Collaboration. Share ideas, solutions and experiences across all silos. Collaborate physically and digitally. Use video, IM, YouTube, Twitter, etc to share and discuss across the entire enterprise. "If Xerox knew what Xerox knows" is a famous quotation from an organization inventing everything but not exploiting anything to their benefit.

Plateau of Productivity? Yes, unless we focus on interfaces and collaboration instead of steering and controlling Enterprise Architects have definitely reached the Plateau of Productivity. What do you think are the constraints for enterprises making more usage of Enterprise Architects to avoid the Success Trap of staying in the Exploitation domain and getting stuck?

Organizational rust and cholesterol

Organizational rust and cholesterol develops in every successful organization over time. In the book "In search for excellence" Peters and Waterman identified the 43 most successful companies in the US in 1980. Out of those 48 companies that were outperforming all other US companies only three would make the list today. What happened to the others?

It's an easy question to ask and much harder to answer. The background issues can quite easily be identified, the cure is much harder to prescribe.
A successful organization develops over time one or more of the following:

• Misuse of resources. The once successful departments get the majority of resources out of old habits.
• Misaligned strategy. What was once a great strategy and market position is today just stubborn actions.
• Leadership dissonance. Blame and backstabbing is spreading throughout the leadership team on different levels.
• Organizational immobility. Misalignment and inability within the different organizational units to move forward. It's one step here and one step there but never synchronized and at the right time.

How can this situation be fixed or even avoided?

In a recent HBR article on the theme "Change for Change's Sake" the authors argument that organizations should practice small and larger regular changes to practice the change organizational muscle. They propose a questionnaire that can easily help identify when it is time to initiate a small or larger change in order to avoid rust and cholesterol. 

An interesting thought is to measure the number of ongoing changes in an organization. How well are we at actually change? Do we have an ongoing readiness for change? What different types of changes do we have ready to apply if and when we need them?

What level of rust do you have in your organization and which changes of different size can you make use of?

How to manage several business models at once?

Facing stiff competition from new business models, regardless if it's new or incumbent organizations, is a challenge for any company. Customer loyalty is jeopardized, profits and cost structures are under at risk and careers can go overboard. Developing new business models is a challenge of its own, but how should my organization respond to the challenge of a disrupting business model in conjunction with our existing way of doing business?


Netflix is an excellent example of a disruptive business model. It works in two ways; first it operates over the internet avoiding fixed costs and local administration, secondly it uses the collective intelligence of its customers to make recommendations based on past rentals. Combined Netflix offers their customers simple and easy movie rentals with much higher viewing experience. How to combat that for an existing video rental business?
Should we:

• Focus on the segment that wants local video rentals taking the risk that they eventually will switch over to the Netflix model?
• Setting up a new business that competes head-on with Netflix taking the risk that we never will be able to do it better than Netflix?
• Creating a completely new business that disrupts Netflix and thus puts us ahead of Netflix taking the risk that our current business will cease to exist?

The text book guide is to either focus on a smaller segments that are not interested in the Netflix model (Porter) or setup a completely separate business with its own value-chain attacking the disruptor (Christensen). 

My take on this problem is to analyse the basic need that the current business model caters for. Can we continue serving these customers so well that they will never switch? If that is the case then change the existing business into clearly market that value and let the other customers leave for the disruptor. Then we have a healthy business that clearly provides a stable customer value that will continue to be there. If not doing this radical step quickly you risk loosing so much money and credibility that there is no business worth restructuring.

Then launch a series of separate businesses as experimentation. You don't know if the disruptor will eventually succeed or if it will create new opportunity and need for other business models. These experiments should both compete head-on and disrupt the disruptor. What's important is that each of these businesses operate separately from the others with its own culture and value-chain where it is necessary according to the business model (customer value, revenue model, cost structure, partnerships and profits). Cooperate with the rest of the organization realizing economy-of-scale where the business model does not mandate it. 

When facing disruptive business models one must be clear that the disruptor is starting small and moves fast. When the disruptor becomes large enough to threaten your existing business it is generally too late. The best prescription against this is to continuously run business experiments in small scale in different locations involving parts of the organization. Then you will try out market opportunities, find better cost structures and be ready for disruption yourself.

Inspiration for the post is from MIT Sloan Management Review article What to do against disruptive business models?

IT driven innovation

IT becomes a major driver behind increased speed and results from innovation. With manufacturing becoming a commodity and marketing increasingly hard and expensive many organizations are turning to innovation as the new revenue source. Innovation can create higher valued products from existing manufacturing capacity and simplifies marketing through improved experience.

IT greatly speeds up innovation and increases the effect in four areas:

• Measurement. How do you measure the current state of affairs?
• Experimentation. What is the value of this idea?
• Sharing. Who can benefit from my new knowledge?
• Replication. How to scale the new innovation throughout the organization?

These areas are important as separate activities and most organizations do one or several of them. They reinforce each other further magnifying the effect from each area. Improving measurements gives you better data easier and cheaper. Experimentation thrives on faster and cheaper measurements. Sharing those results further increases the value across the organization. Replicating successful business innovation enables faster scale ups and shorter time to market.

Measurements is radically improving the way we collect data and sort it. Nano data from click streams, emails, discussion boards, web trends, CRM, supply-chain, etc. Business intelligence is tapping into this area; needs to tap into the very bits and pieces of the nano data and make it available to the organization.

Experimentation through IT tools based on improved measurements. Measurement both to feed into the idea generation and to evaluate the effect from different ideas. Retailers can use IT to efficiently instruct different displays and then use improved nano measurements to follow up sales and profitability. New business models can use IT to track the consumption of goods and services through RFID tags on individual levels.

Sharing the results of experiments and the measures improves others to innovate. They don't have to invent the wheel again, can get the exact data measure setup, sharing ideas that for some reason failed (sometimes the most important aspect as failures seldom gets documented and less often shared).

Replication of innovations once they've been evaluated and proven is the key to successfully reap the benefits. The first three areas help organizations find and nurture new innovations. IT makes it possible to take that innovation and roll it out quickly and seamlessly. IT drives replication of business processes themselves. An example is how a bank is replicating a new loan process across all branches, a new customer return policy, displaying new food products in a retail store. Measurements enable the follow up of how well replicated the innovation is.

Collection of nano-level data is the foundation for IT driven innovation and replicating the innovation across the organization enables fast time to market.