The document discusses the importance of systems of record for businesses. It notes that systems of record are highly structured, transactional, reliable, and core to the business. In contrast, systems of engagement are loosely structured, quick to adapt, conversational, and at the edge of the business. The document advocates developing a strategy to modernize applications and transition to newer architectures like cloud, while ensuring systems of record still meet business needs as engagement systems evolve.
2011 Geoffrey Moore wrote a paper titled “A Sea Change in Enterprise IT” where he talked about the shift underway in many organizations moving from the traditional approach of managing enterprise IT to a much more flexible approach. Moving from a highly structured model where IT started in large organizations (using tools like relational databases, 3rd generation languages and point to point interconnection techniques) to a model where innovation is taking place at the edge of the enterprise, at the user interface, forcing IT organizations to adopt to the new consumer demand, or feel out-of-touch with their users.He separated the underlying IT systems into two groupings:1) Systems of record <READ THE SLIDE>2) Systems of engagement <READ THE SLIDE>Much of the mobile, social and collaborative trends that organizations are coming to grips with are focused on the balance and interaction between these two types of systems.In this presentation I’d like to talk a bit about the on-going importance of systems of record, as well as how their maintenance, management and governance actually enable you to effectively delivery systems of engagement. And increase the delivery of business value
Nearly every business has a few drivers in common. They want to be fast, agile and financially successful.For IT organizations this means overcoming an array of challengesLike the constraints of their existing systems, processes and measurement techniques.The existing systems which for most organizations are Systems of records contain all the detailed information of the organization, but they also consume 80% or more of the budget just to keep them operating. This is a huge constraint on what organizations can do.Also, the existing processes in most organizations were not set up for a mobile workforce that can work wherever and whenever they need to. These processes can drag the organization down with all their built in latency.Finally how performance and value are measured may not have kept up with the current business models -- further confusing the business leaders with performance measures that were meant for another era.Addressing these issues can significantly improve performance for organizations.
Since systems of record have been around since the inception of modern information technology, for most organizations they have built up layer upon layer (a monument to previous success), allowing our budget spend to become calcified justkeeping the lights on. If organizations are not careful (especially those who have been through mergers and acquisitions) systems of record can be stacked climbing to an overlapping tower of functionality that may or may not actually deliver business value. They are maintained, because of inertia, not any specific requirements. Some real discipline and enterprise architecture is required.I was talking with a new leader of a large financial institution a while back and the first thing he did when he came into the organization was state:“for every system where I can’t find an owner, we will turn it off at the end of 90 days”After the cut off date, they turned off 15% of their systems.After they did that, they only had 2 systems where someone sheepishly made a commitment and said – I’ll own it.The rest of the systems were just a drain on their manpower, budget and an anchor on innovation.
I’d like to spend a few minutes talking about what’s changing on the IT side and why our perspective of systems needs to change.<talk about the slide>
There are a number of technical trends are well understood. What I’ve tried to do here is extrapolatethem out 5 years to 2017.You can probably see how unimaginably different these numbers become from what we’re used to todayA good example here is in the healthcare space. Today it costs between $2000 and $5000 to truly capture an individual’s genetic information this will shift to under $100 in 2017. This means that if the processes and systems can keep up, we don’t need to prescribe drugs based on how most people would respond (the way we do today). Instead we can use targeted drugs (that we know work) based on someone’s genetic makeup.This is why effort spent on the transformation of systems of record is so critical – today. Planning for this world where we can have petabytes of storage on mobile devices like cell phones will enable whole new approaches when compared to the most modern cloud based techniques we think of today.
Going forward it will be all about choices – this is one area that I think HP’s approach to cloud and modernization is a real differentiator from others. It is about enabling you to make choices, not forcing our choices on you – but that’s really someone else's presentation.What’s key for this presentation though is that our mental model needs to shift to enabling and “and” not an “or” world for the business, where IT organizations enable selection among possible solutions, rather than between them. A key part of this is maintainingan application portfolio that supports this more flexible view.Where organizations can use information to make business decisions wherever and whenever they need to using services pulled together from a variety of sources.
There will not be one right answer that applies to everyone.Instead almost every organization will need to support a hybrid delivery model based on an organization’s objectives, adoption maturity, risk profile and other factors. Each business will have a unique entry point or “on-ramp” mix that they need to address.What we think of as possible for the future of computing has changed radically, recently. Today we have computing solutions getting driver’s licenses. Systems that can perform pattern recognition and derive the context of spoken and written text are becoming common. The level of automation that is possible is quite different than just a few short years ago.Organizations will take many of the concepts they have learned deploying and automating the information technology processes on their path to cloud and apply them into other parts of the business. Radically reducingthe latency in the business decision making process.I think of Cloud Computing as an example of the degree of automation that can happen elsewhere in the business. The systems of record provide that solid foundation for automation. Without a stable basis for pattern recognition, workflow, simulation and modeling, business automation will be a difficult goal to achieve.
Now is usually when people start asking how do I get started on preparing for the future?You have to create a strategyYou need to understand your current environment. Perform some sort of current situation analysisDefine an architecture plan (start small --defining meaningful projects after all if they are not impactful, no one will care about the results)Prototype areas of that are unknownShutdown low value applicationsSeparate out normal from the anomalies for the business and begin to automate “normal” and focus the people where their creativity is needed.
Planning is critical for having these efforts succeed as well as the organizational change management to ensure the whole organization is ready for the change can be the difference between success and failure.So a structure with that delivers work products that can provide a level foundation of understanding is important.HP also has personnel and processes to assist in the assessment of a businesses application portfolio.As well as personnel, processes and tools to facilitate the modernization of applications. Our approach classifies applications into 4 domains for architecture life-cycle planning:
Some applications may remain in their current environment unchanged, but Applications that need work are categorized four domains are:Re-Host: This is the most non-invasive approach and is for the most part a lift and shift. Migrating the application as—is over to a new infrastructure that would continue to be managed in a traditional fashion.Replace: With some applications it may not make sense to continue maintaining them. Alternative replacement or condensing strategies may be possible. For example – if our client has a custom (or definitely if it has many) CRM application, it may make more sense to replace this with a cloud based CRM cloud offering. Such alternatives may provide more features and benefits at a more competitive total price.Integrate: The integration of off-premise cloud applications is frequently identified in analyst reports as a barrier to cloud adoption. For this reason, our approach may include the build of our clients enterprise integration platform enabled via an enterprise service bus, or it may require the wrapping of existing legacy applications to expose them as web based services. This domain would leverage our SOA and Integration Services.Re-Architect / Re-Factor: These two options are the most intrusive, but may be required for applications written in a legacy language (COBOL, IMS, PL/1, etc.). Cloud techniques provide highly saleable and parallel processing enabling capabilities that if organizations really need to take advantage of these features, significant redesign will be required. Although many people talk about cloud bursting as a way to add greater flexibility, actually enabling applications to do this is much more difficult than just talking about it.It is getting to the point where I believe the word application is becoming meaningless, since it is more of an aggregation of interconnected services than what is normally thought of as a stand along application.