Enterprise architecture and digital business

Allister-Homes-Profile-pic---small

 

Allister Homes
Senior Systems Architect
University of Lincoln

Gartner EA Summit Day 1

There are attendees at the summit from all over Europe and it’s probably fair to say that as someone from the HE sector, our EA practice is somewhat more modest than a lot of others represented here; I think that would be the case for almost any HE institution that came here, I don’t mean just Lincoln.

I haven’t seen any information on the number of delegates, but I’d estimate there somewhere in the region of 400-600 and it’s quite a busy summit in terms of the number of people and the number of sessions. I thought the most straightforward way of blogging this would simply be to summarise each session I attended, so here goes.

Architecting the digital business ecosystem
ecosystemsAs Gartner’s opening keynote session, this started by concentrating on the fact that more and more businesses operate in the digital world – not just having a presence, but offering their services and interaction digitally. Garner’s figures are that 22% of businesses operate in the digital world now, and by 2020 that figure will be 83%. Digital business will mean business growth, and those who do not embrace it will lose out to those who do. We heard about the seismic shift happening, caused by the Internet of Things and what Gartner calls the ‘Nexus of forces’.

It is estimated that by 2020 the number of Things connected to the internet will outnumber people by at least five to one. We heard a story of how a connected home might, in a few years’ time, react to a tree crashing into the bathroom during a storm whilst the owner is away by cutting off electricity and water to protect the property, contact the insurance company and start arranging quotes, pass the information to the police and ensure the property is safe and secure. As Enterprise Architects we need to be aware of new technology and innovations so that we can become ‘vanguards’ and shape its use in the enterprise, which will become continually more consumer focused rather than enterprise user focused.

We analyse too much and synthesise too little

This session was focused on trying to change the way EAs think. Rather than always relying on analysis, the suggestion was that we rely on synthesis a little more. We were told how analysis does not work well when there is no data, there is ambiguity and when there are dilemmas, and then a short history of synthesis starting with dialectical reasoning. Some of the simpler examples used where synthesis can were how to cut costs given a particular banking scenario (don’t rely on the distinct possibilities provided by analysis) and the ‘chicken or beef’ question on aeroplane flights (either ask for both or eat before boarding!).

The state and innovation: from fixing markets to market making

state and innovationProfessor Mariana Mazzucato from the University of Sussex presented on innovation led growth and the role of the public vs private sector. She described the cartoon image many have, which assumes all innovation happens in the private sector whereas public sector is rational, boring, has dinosaur bureaucrats and exists (in terms of innovation in markets) to simply level the playing field. However, she went on to explain that science is simply not funded enough without state support, which is needed to fix market problems. In fact, public sector funded innovation is often mission oriented, such as sending mankind to the moon, and massive public sector innovation and investment has led to much of the technology that makes things like smart ‘phones possible – think of GPS, voice recognition, touch screens, cellular communication, signal compression and more.

What can sometimes be forgotten in the public sector, though, is to apply what is taught and used elsewhere, such as the approaches taught in business schools and lessons that can be learnt from great innovators. One particular example highlighted was that of higher investment risk green opportunities which are starved of private funding. In Germany in the early 2000s, when private banks in Europe were reducing loans and the economic crises was in its early stages, the public sector in Germany provided substantial mission-oriented funding for environmental and climate protection innovation.

Application architecture for digital business

This session, delivered by a Gartner analyst, concentrated on new approaches to application architecture for delivering business digitally. It was emphasised that the focus should first be on business outcomes which are then converted to technical goals, which lead to technical characteristics, then principles and then patterns. Most organisations are still using three tier (or n-tier) client/server architecture, which is typically designed with static architecture and linear flow in mind. The approach does not work so well with consumers using various devices and with business that needs to change rapidly, and so an application architecture of apps and services was suggested instead. This takes service-oriented architecture several steps further, and encourages the use of individual apps for particular personas, devices and scenarios rather than one large client application to do everything, uses services to support the apps, and encourages a many-to-many relationship between apps and services. In this scenario services are broken down much more than they typically are in most environments today, becoming microservices for web-scale SOA. Examples were provided of how the likes of Netflix, Facebook and Amazon use the microservices to scale to massive numbers of concurrent users and with the agility to make hundreds of updates per day to production systems by having these very specific microservices which are independently deployable and scalable. The difficulty is that although they provide separation of concerns and are individually easy to understand, you end up with a LOT of services, and have to radically alter the rest of the architecture because there is often no centralised application or data. Third normal form is out the window with this architecture, and a lot of what was done as server code becomes client code again.

Nissan Group: using enterprise architecture to enhance efficiencies and business growth

Enterprise Architects from Nissan presented on their use of MEGA at the Sunderland plant. Nissan had a lot of legacy applications and infrastructure at Sunderland, but not necessarily a good corporate understanding of all of it. Three main drivers led to the need to change this: a security incident, the transfer of some service to third parties, and a corporate objective from Japan to understand and standardise business processes. A project was launched to document 500 applications, the central technology infrastructure and the business processes, and to link them all together into a single EA model. A team of five full-time staff were set the task of doing it, and although timescales turned out to be a little too ambitious, much of the work is now done including the development of an application portfolio outside of the MEGA model and the creation of a bespoke metamodel.

The cloud computing scenario: examining cloud maturity and adoption in the age of digital business

This session looked at the adoption of cloud by businesses, how to make assessments and what to consider with a cloud strategy. Gartner’s analyst explained that cloud delivery options are becoming more varied and complex, leading to a “spectrum of cloudiness” that EAs need to understand in order to make the right decisions for the business. It’s not just the delivery model that needs to be considered (public cloud, community cloud, virtual private cloud and private cloud) but also the execution model, for example whether hardware and software is isolated or not. Cloud offerings are still changing quickly, for example by making improvements to security models, and although maturity is growing it is too early to put a final cloud strategy in place; a strategy is needed, but it will need to keep pace with the constantly maturing cloud technologies and offerings. Vendors sometimes complicate this, and an EA needs to be able to break through the vendor ‘fog and hype’ to understand what is really being offered.

It was emphasised that whether we like it or not, many (not all) vendors are now going cloud first with new solutions rather than traditional software first, which means the decision is shifting from ‘whether and when to go cloud’ to ‘how and how much’ to go cloud. The reasons for moving to cloud solutions is not always cost-based; there is value from agility, the offloading of responsibilities, business agility and the advantages that provider investments provides in terms of scale and security. An interesting element of the presentation was of how business users tend to focus on benefits of cloud such as speed of change (sometimes bypassing IT departments) and that IT and EA focus tends to be on ubiquity, scale and so on; there needs to be a balance and a meeting point between the two views.

In reality most organisations will use a hybrid of cloud services, and the cloud strategy needs to consider the “spectrum of cloudiness”. Comment was also made that not all vendor offerings are created equal, and work must go in to understanding the differences beneath the surface. There are some large vendors, such as IBM, who offer cloud services and will also build cloud solutions, and will often mix both in the same conversation which can lead to confusion and complex, bespoke solutions. The five questions the presenter suggested be asked when defining a cloud strategy are:

  • Where and how should the enterprise consume cloud computing services?
  • Where and how should the enterprise implement private cloud environments?
  • How will we secure, manage and govern cloud services across internal, external and hybrid environments?
  • How does cloud computing factor into our application strategy and architecture?
  • Are there opportunities for the business to become a cloud computing service provider?

Selling EA with stories: start with the end in mind

selling eaEE, the mobile telecoms company, presented this session. EE has a mature EA practice which is engaged with all levels in the organisation. It recently refocused EA into three areas:

  • Technology that underpins enterprise
  • Business capability
  • Business architecture

It has a comprehensive EA knowledge base or model, using Sparx, and has a philosophy of ‘doing just enough, just in time’ because otherwise all the time is spent trying (and failing) to keep everything perfectly finished and fully up to date instead of spending time innovating, working with business users and influencing change, which is where the real EA value is.

An opportunity arose for the EA team to create a vision, sell it to the business and propose a set of new initiatives. The team achieved this by first creating a compelling vision. The vision was rooted in the present day and based on information already held in the knowledgebase (architecture model), and focussed on the main business problems and desired business outcomes. The vision was communicated through personas and stories, which were designed to be both factual and appeal to emotions, by highlighting key frustrations and weaknesses and how the vision would help overcome them. The vision, including stories, was presented on one single sheet, including short biographies of personas, themes in columns and the use of icons and short stories. No formal architecture documents were given to the business at this stage, but the detail had to be ready if the outcomes were positive and the team was invited back to the senior stakeholders for further dialogue on roadmaps, sequence of capabilities, business benefits and costs. The approach was successful with three of five major initiatives being started.

One thought on “Enterprise architecture and digital business

Leave a Reply

Your email address will not be published. Required fields are marked *