Bridging the GAP: Unified Approaches to Governance, Architecture and Procurement
Jonathan Murray is VP and CTO of Microsoft for Europe, Middle East, and Africa. This is a new role for the corporation. He runs a group of 30 technology officers around the world that work with government, educators, and others on policy implications of technology.
The World Bank today spends about $1 billion a year on ICT related projects. Sixty percent of those projects have been failing, which has major implications on emerging markets. Jonathan had worked as a technology officer at ARCO before coming to Microsoft, so had a background in enterprise scale IT issues. An example is Siemens, which has 350,000 desktops in 120 countries – highly complex environments. At Microsoft he put in measures to gather data on customer satisfaction with their services. When they started that process they were surprised by a couple of things – one was the larger the customer was the less satisfied they were with Microsoft. So they then formed a group to manage those set of customers separately from the rest of Microsoft business.
Out of that work came some interesting insights. One thing they learned is that all those companies were mired in complexity. What are some of the core issues from those complexities, and what are some of the best practices that come out of those companies?
They went and looked at the Educause data on most important issues, and were struck by the parallels with the private sector. The issue of funding for IT is high in both areas, as are security and identity management, and strategic planning. The issue of ERP implementation has largely been dealt with already in most large enterprises. In private enterprise security is still important but has become part of the system.
He’s surprised by how low the issue of Governance, Organization, and Leadership turned up on the latest Educause survey – in private sector companies it ranks much higher as an issue – governance is the glue that makes this all work.
He puts the issues into a taxonomy of Governance, Architecture, and Procurement.
Governance includes strategic planning for IT, faculty development support and training, Governance organization and leadership for IT.
Architecture includes security and ID management, Admin and ERP systems, infrastructure management for IT, elearning, portalks, web services
Procurement includes funding, strategic planning, and support and training. How do you make sure the procurement cycles map to the rapid development of technology? This is a huge issue for governments.
In plain terms:
– in governance you deal with competing needs of many diverse stakeholders (how do you satisfy demand while keeping control of stability?), more demans than capacity, everyone is an IT expert
– Architecture – unless you have good architecture that is well managed and has a long-term plan it is very difficult to plan for an manage the complexity.
– Procurement – the budget process.
The GAP principles – derived from the best in class companies in how they manage IT.
– Economic pressures of the early 200s lead leading companies to develop a set of best practices in IT governance, architecture and procurement: The GAP principles.
– Many higher ed institutions continue to implement GAP approaches which leading companies evolved away from in the late ’90s
– Higher edu adotpion of the GAP principles would enable accelerated deployment of new services and technology
What are the principles?
1. IT is a service provider to the business. – Buiness units and information technology organizations need to be intimately linked through managed engagement processes.
2. The CIO requires real authority – CIOs need effective authority to mandate architecture standards across organizational boundaries. This is absolutely critical. The best of class companies have CIOs with the authority to set and enforce standards. This is a thin layer of authority which runs horizontally across the entire organization.
Architecture – Good architecture demands abstraction. Too much of what we build is tightly vertically coupled. Increasingly what we need to do is build flexibility built on open standards.
Procurement – Architecture is the foundation – a long term strategic model is required for core architecture procurement. Service orientation in architecture enables flexibility. SLAs are ok, but they tell the service provider the minimum they need to know, so they’re not the end-all.
Over the history of enterprise computing, nothing has gotten simpler over time. The late 90s we saw the IT “abyss”, with IT spending growing rapidly (no old stuff was being switched off), operations and maintenance dominating IT budgets (that’s still the case today – about 60% of budgets are spent on this, with only 28% being spent on adding new value), the complexity of distributed computing environment was exploding, and new development was ineffective. (1997 McKinsey research study).
Three catalyst events:
– Remediation of systems for Y2K – this was the first time organizations understood what IT systems they actually had, because they had to inventory.
– Stock market crash and bursting of the internet bubble
– september 11 attacks – told companies that the world is not predictable, which had a big affect on the trend towards outsourcing.
What do leading private sector organizations look like today?
– the web as the fundamental fabric – open stanadards, xml, http, etc. internal as well as external
– application abstraction through service orientation. So need to know what the core services are that need to be developed and then delivering them in abstract ways
– Systems abstraction through virtualization – fully virtualized data centers. This is very big in the financial services industry. Compute, disk, memory, IO are just a pool of resources in the data center, which can be allocated on demand.
– Federal models. There is a CIO empowered to set standards for infrastructure that have to be implemented company wide. Includes desktops, middleware, and service infrastructure. CIO has the board’s authority to mandate that all development will use these standards. Having account managers, sitting day-to-day inside the client units is a best practice – making the standards understandable to the clients and feeding back the information from clients on business needs to the IT organization.
– Architectural “hegemony” – Shel points out that in research institutions you need to plan for those disruptive technologies that will arise and break the architecture.
– strategic partnership – picking vendors whose products are committed to the set of standards that are being implemented.
– The rise of shared service models – how do you make systems and services work together across units to service the customers. One example is in South Africa where they are trying to combine individual institutions into a system, and want to offer common services.
Benefits of good architecture – not about specific products, but about standards.
– Abstraction of complexity – making sure that applications all go through the same middleware layer, for example, to allow changes in the back end without application changes.
I asked about the conflict between abstraction and ERP implementations, and Jonathan replied that most well managed companies implement ERP in order to enforce standardization of business policies and practices in order to get a global view of their business, and use the ERP as a transaction engine, and then build a series of data warehouses that people interact with to get information. Not everybody gets access to the ERP, and nobody gets custom reports. At Microsoft the data warehouse applications provide HR data, for example that’s no more than 20 minutes old.