Category Archives: Cloud

There is no “I” in IT: Oh wait yes there is.

The Dynamic Enterprise
I’m in the airport in Victoria on my way back from giving a talk sponsored by the Office of the CIO in the Government of British Columbia entitled “Enterprise-Scale Business Transformation with SOA and BPM”. Increasingly, I’m seeing efforts around large-scale Business Transformation involving the dynamic system of Business and IT.

In my talk, I asked the following question:
How do you achieve “Business Transformation” in a system that:

* Is exceedingly interdependent (or in the words of the CIO here, “tangled”)
* Responds unpredictably and dynamically to change
* Is complex enough that no single human is capable of understanding it all

In short, a system that exhibits dynamic complexity as well as high-order entropy. I have an elegant solution for this problem, but this blog entry is too short to contain it.

One Continuous Mistake
I’ve seen lately many methodologies that depend on “objective data” and measurement to align organization. Don’t get me wrong, measuring the business value of any activity is a way to align activity in a paradigm of continuous improvement. I’ve long been an admirer of Toyota Kaizen, whose principles of long-term employment, respect for the individual, empowerment and “Moving Forward” (continuous improvement) are based on a balanced and “lean” adaptation philosophy that evolves dynamically with the business environment. Not only has this facilitated the transformation of Toyota from a car maker to a hybrid car maker, but it powered their transformation from a maker of mechanical looms to a carmaker after World War II.

The granularity of the improvement cycle is a key success factor in using continuous improvement as a paradigm for business transformation. This is because a short cycle reduces the risk expose to certainties that “simply aren’t so”. @bmichelson @mtdonahue tweeted it very well by saying:

Walking on water and developing software from a specification are easy if both are frozen – Edward V Berard

What is of higher value in transformation of dynamic systems is not how you succeed, but how you fail. The ability to fail continuously and improve continuously is the secret to transformation. This requires the organization to firstly overcome fear of being “locked in” to static business processes and governance policies. Unlike the old infrastructure, the new infrastructure of Business Transformation is based on continuous improvement on short cycles and thus empowers active participation and collaboration in the shaping of optimal process. As Toyota knows, nobody knows more about process optimization than those who execute the process.

Mechanisms need to be in place for lowering the cost of failure. Short cycle times address this (borrowing from Agile), but also things like Cloud Computing’s Elastic payment model dramatically reduce the cost of failure. The era of huge IT failures should be replaced by the era of distributed intelligent transformation. Thousands of microscopic iterations of process improvement managed within a frame of trust and coordination.

An airplane is continuously off course, but also continuously reorients towards its best sense of where it’s going. As such it is capable of optimizing itself both locally (to avoid weather features and air traffic) and globally (to reduce fuel consumption and travel time). This dynamic interplay of local and global optimization is in the sweet spot for the transformational agenda.

There is no “I” in IT…oh wait yes there is.
I’ve had the great pleasure to read @davegray and his insightful post about a general theory of Information Relativity

The lovely insight here is the fundamental lack of separation between information and intention. I keep seeing transformation methodologies that depend of “objective information” to align organization, but this insight helps us to understand that information (the “difference that makes a difference”) is different from data and that the transformational power of information is inextricably intertwined with principles of presenter bias, observer bias and other key factors.

A nice commentary by

Nowhere is this more evident in the emergent behavior of “KPI Gaming” that manifests in organizations who have successfully implemented organizational alignment through continuous measurement of Key Performance Indicators, and have tied those KPIs to individual and organizational incentives, job descriptions and roles and organizational structure. I spoke with a CIO in Australia who bragged about how dramatically he improved project delivery time to get a fat quarterly bonus, but received a wink and a “no comment” when I pressed him on whether he rescoped any projects in order to get there. Obviously if you cut one project into 10 you can deliver projects “ten times faster”.

This connects with the idea that “direct observation” of reality is rare and almost always mediated through our past experience and the filter of social expectation, culture and near-instantaneous mental reinterpretation. Mediation is a fundamental pattern that underlies scale in organization, as it provides a way to enforce policies, culture, route information and detect patterns. Routing information means that the organization gains the ability to amplify inspiration and best practices, while policy enables organization to enforce best practice and define itself. Of course the “mediation” pattern in Service Oriented Architecture typically involves placing a piece of hardware and/or software in the path of message flow, but there are technical architectures that perform mediation functions embedded in the endpoints.

For humans, embedding organizational culture, policy, pattern detection and information routing in the endpoint means fostering a culture that understands deeply when to escalate issues, when to share information, where to share information and when to report, stop or perpetuate behaviors. This internalization of behavioral patterns, habits and culture allows organization to achieve what Peter Drucker refers to as the goal of organization: “To enable ordinary people to do extraordinary things.”

Lorraine Lawson wrote in her blog today:
“Honestly, I’d rather write about anything besides governance, particularly SOA governance.”

This post explores the concept of Federation, the balance of regional and central authority as the differentiation between Governance and Management. Ultimately this brings me to Daryl Plummer’s folk definition of Federation which is:

“What do I have to give up in order to be a part of something bigger”.

Linking that with Drucker’s “organization enables ordinary people to do extraordinary things”, one is led to the hope that scale does not equate to life-sapping and stultifying constraint and that the autonomy that one yields to belonging is amply compensated by the scale and longevity (legacy in the positive sense of the word) of Enterprise. Although both Dilbert and “The Office” suggest that the sacrifice is your life in exchange for 20 pieces of silver, we must envision a world where great problems can be addressed at great scale.

Cycle time for continuous improvement empowers the individual to make changes in their local environment and thus experience empowerment and ownership. It’s only the rigidity of policy and process that leads to the snuffing of the individual life-force in the Enterprise.

Or perhaps Mother Teresa is right and we cant do any great things, we can only do small things with great love.

Posted in Cloud | Tagged , , , , , , | Comments Off on There is no “I” in IT: Oh wait yes there is.

Inside of a Dog, it’s too Dark to Read: SOA and Business Variation

Groucho Marx once famously said “outside of a dog, a book is man’s best friend. Inside of a dog, it’s too dark to read”

The amusement is that the word “outside” is semantically ambiguous. What’s less funny is the SOA Manifesto stating that they will follow the principle of “Pursue uniformity on the outside while allowing diversity on the inside.” Much like the Groucho Marx quote, the antecedent for “outside” is ambiguous. Outside of what?!

(Picture of Groucho Marx looking a bit like Ron Schmelzer)

Why is this statement problematic?

First and foremost, I chastise this statement because it is sloppy. Sloppy architecture is what got us into this mess in the first place. But more importantly, this statement reflects a narrow, IT centric view of the scope of Architecture which leads down a dangerous path. Complexity theory has a term called a “holon”. A holon is something that appears to be a whole–yet is a part of something larger (e.g an organism is part of a biosphere). By confusing the antecedent for “outside”, the SOA Manifesto focuses on the holon of the service interface. If so, the reworded principle would be “Pursue uniformity on the outside (of the service interface) while allowing diversity on the inside (of the service interface).” The problem with this statement is that it ignores the concept of diversity on the outside (of the service interface).

Diversity on the outside solves the business complexity problem
Lest you think I’m splitting dog hairs here, why is this of any importance?

Because everyone is presently overfocused on the complexity of the IT Supply rather than the complexity of Demand on IT. The business is driven to compete, innovate and requires Internet style services that scale and can be consumed on any device in any language, anywhere and any time. They demand self-service, personalized, localized, mass-customized and differentiated customer experiences, and they require access to the same information whether it’s in a customer contact center, on a web browser, an internal application, on a mobile device or at a branch office. Business demands complex consumption patterns on the outside of the service interface.

By ignoring the consumption pattern diversity, attention is only drawn to the IT problem of managing complex IT supply. One of the biggest flaws in SOA implementation, and arguably the reason why Anne Thomas Manes declared “SOA is Dead” is the lack of connection between SOA and the business solution. The ability to dynamically compose solutions into composite applications, mash-ups or business processes using business services is the central underpinning for agility and business transformation.

Managing Business AND IT Complexity
This requires the ability to solve the problems of Business Complexity and IT Complexity. Services are just tools that allow you to trade consumer complexity for trust, thereby improving manageability. Processes are tools that take full responsibility for complexity, thereby enabling optimization. The primary function of Services in this context is to manage the interaction between two specialized organizational subgroups. This allows one group such as business unit to consume the services of another group say, IT without having to know the gory details of the implementation of the service. Insulating the business from the complexity of IT is helpful, since the IT subgroup is highly specialized and therefore hard to understand. However, two implications arise, firstly that the business group is also highly specialized and therefore complex. Secondly that IT cannot abdicate its responsibility for IT Process, unless of course it is prepared to outsource the service. The provider view of a service is process-oriented, only the consumer view of service is service-oriented.

The “new normal” Business Environment
A good example of this complex consumption pattern is Software AG’s recent partnership with Tom Tom Work–who are now able to offer fleet management as-a-service to their business customers. The ability to offer outsourced transportation and logistics management enables Tom Tom to provide a new differentiated service on top of the core competencies of their organization, namely location-based services. The power to dynamically recombine core capabilities into new product and service offerings while simultaneously offloading non-core competencies to partners reflects a “new normal” business environment. Complexity should be seen as a function of specialization, and organizations should identify which sources of complexity they want to embrace through process-orientation, and which sources of complexity they want to eschew through service orientation. Eschewing complexity through service-orientation requires the discipline of segmenting specialized Enterprise subgroups and actively deciding whether your organization wants to invest in world-class processes or divest services to outsourcers. No single person in the Enterprise can understand the total complexity of the value chain, but this does not lead to weakness as long as you can align internal and external subgroups through trust or by developing a trusted framework of operation.

Trusted Frameworks

A trusted framework allows you to manage complexity housed in many discrete silos without requiring explicit trust between specific providers and consumers. By externalizing trust into policies that are uniformly enforced, we create a trusted environment. An example is automobile traffic. Automobile traffic is exceedingly complex, and no single driver can understand what is in the minds of all the other drivers. In fact, when one goes on the road, it is not because one trusts each and every driver explicitly, rather that there are enforceable rules of the road that have been declared and enforced. Individual trust relationships between drivers is not scalable. So a driver trusts in the framework of policy enforcement. Similarly, trusted environments hold the key to enabling spontaneous networks of collaboration between Enterprise subunits and external partners and suppliers.

As such, the business and IT functions need to develop a framework of trust that creates the possibility of continuous improvement–an evolutionary paradigm that at the Enterprise scale requires continuous measurement, continuous optimization, continuous alignment and continuous collaboration. If SOA and Enterprise IT are going to bury their heads on the “inside of a dog”, it will be much too dark to read the intentions of the business.

My 2 cents,

Posted in Cloud | Tagged , , , | Comments Off on Inside of a Dog, it’s too Dark to Read: SOA and Business Variation

The Legacy of Enterprise 2.0–Concluding thoughts from the Conference

The absolute highlight of last week’s Enterprise 2.0 conference was meeting in person and online many very bright people including Nenshad Bardoliwalla, Susan Scrupski, Michael Krigsman and many others. There’s certainly a strong discourse here about advancing the agenda for Enterprise computing.

As with many advanced topics in Enterprise computing, it’s very easy to take a potshot: as we all know:


* SOA is Dead (Anne Thomas Manes)
* Enterprise 2.0 what a Crock (Dennis Howlett) and
* Cloud is Water Vapor (Larry Ellison)

One link that got me to thinking was posted by Nenshad–Tomio Geron from the Wall Street Journal blogged about the vast number of venture backed startups that have failed this year.

Pundits aside, I think there is a very easy “Crock Test” that can definitively answer Dennis Howlett’s question. The question wont be answered by a failed panel discussion. No matter how pithy, it wont be answered by a blog post. The closest thing to the “Crock Test” came from Susan Scrupski’s “last and only response to crock-gate”, a list of organizations involved in the Enterprise 2.0 Adoption Council. The problem with the list is that the names are just names–you kind of want to be able to click on them to see the case studies, if any.

Which brings me to the “Crock Test”. Unfortunately, the test will require some degree of patience. We will definitively know if Enterprise 2.0 is a crock if in 20, 30 or 40 years we can look back on all of the Enterprise 2.0 legacy software that has been created.

As I mentioned in 5 definitions towards the maturing of Enterprise 2.0, legacy is another word for the software projects that worked. So here’s hoping we will be looking back at Susan’s list of companies and seeing a list of case studies about the depth and scope of transformation and a list of the huge successful companies built on top of Enterprise 2.0–not a list of failed ones.

To be fair, the failed companies on the list come from a large number of sectors–it’s just a sobering list, not a condemnation of #e20

My 2 cents,

Posted in Cloud | Tagged , , , , , , | Comments Off on The Legacy of Enterprise 2.0–Concluding thoughts from the Conference

The Enterprise 2.0 Crock

I think it’s great to have a panel called “Is Enterprise 2.0 a Crock?” at the Enterprise 2.0 conference. Much like SOA conference panels addressing “Is SOA Dead?”


Dennis Howlett posed this question in his blog, entitled Enterprise 2.0: What a Crock. The summary paragraph is quoted here:

Like it or not, large enterprises – the big name brands – have to work in structures and hierarchies that most E2.0 mavens ridicule but can’t come up with alternatives that make any sort of corporate sense. Therein lies the Big Lie. Enterprise 2.0 pre-supposes that you can upend hierarchies for the benefit of all. Yet none of that thinking has a credible use case you can generalize back to business types – except: knowledge based businesses such as legal, accounting, architects etc. Even then – where are the use cases? I’d like to know. In the meantime, don’t be surprised by the ‘fail’ lists that Mike Krigsman will undoubtedly trot out – that’s easy. In the meantime, can someone explain to me the problem Enterprise 2.0 is trying to solve?

Now don’t get me wrong, I believe there are legitimate technologies that can help organizations collaborate and gain significant advantages. But the Enterprise 2.0 crowd for the most part do not show the kind of sophistication about the Enterprise that is needed. OF course there are many many counterexamples of people who are very sophistcated, @ITSinsider @Nenshad @PhilWW @Mastermark and others.

But when I hear panelists say things like:

* The way that business is organized is fundamentally changing, period
* we are breaking down silos
* I don’t know who goes into offices anymore
* Email “reply to all” costs $250,000

I really wonder if there’s enough healthy debate on this topic. It would be nice to see @dahowlett on the panel.

The concept that Silos are something that can be “broken down” shows the ignorance of the people saying it. I wont point a finger at the panelist who used the phrase, and I hear people saying this throughout the conference. But this is a completely wrong concept, and dangerous.

Silos, both organizational and technological are an emergent property of Enterprise. As I defined earlier in the week, the Enterprise can be defined as an organization whose mission requires longevity, size and growth. Longevity creates technology silos, growth creates organizational silos.

Christopher Allen cites Robin Dunbar very well in this post about the famous “Dunbar Number.” We are hitting fundamental limits to human social scalability in the Enterprise.

Until Enterprise 2.0 folks gain a deeper understanding of the day to day reality of the Enterprise, this will continue to have a superficial impact on the Enterprise. If we look back at Enterprise 2.0 in 20 years and can see lots of Enterprise 2.0 “legacy applications”, we can consider this effort to have been a success.

Posted in Cloud | Tagged , , , , , , , | Comments Off on The Enterprise 2.0 Crock