Christian Verstraete at a glance:
- 20 years specialising in IT supporting business processes
- Studied engineering at Université catholique de Louvain, Belgium
- Co-authored Strategic Value Creation through Collaborative Supplier Relationship Management, (2005)
Many IT departments are struggling. It see its budgets frozen at best, while the business requests more and more from them. The combination of cloud, mobility, big data and social is transforming the way businesses interact with customers, employees, partners and suppliers. This is increasingly referred to as the digital enterprise. Knowing most IT departments still spend 60 to 70% of its budget in operating its environments, the need to take a clear look at every pound spent seems obvious.
Indeed, while requirements pile up, while business teams bypass IT and consume services from external suppliers, often without taking a hard look at security and compliance implications, IT is expected to deliver within its budget envelope. Additionally, the business expects IT to be more agile, more responsive, to adapt new technologies (why aren’t my iPhone and iPad supported?) and to develop new functionality faster.
The fundamental question is, how you do that without additional budget? By saving money in existing operations. This is where a lean approach comes in.
Review and optimise the IT processes
The first thing to do is to review current operations. Can we eliminate waste? This requires visibility and understanding of what happens. You cannot walk along your IT supply chain in the same way you do in your
factory, so you need to find other ways to visualise what is truly happening. For that you need a frame of reference and there are two you can use.
The first one is ITIL (Information Technology Infrastructure Library) or ITSM (IT Service Management). ITIL starts from the services delivered to business and recognises five areas:
- Service strategy
- Service design
- Service transition
- Service operations
- Continual service improvement
ITIL is widely used in the industry, although it is sometimes seen as too rigid for cloud-based operations.
The second one is less known but quite intriguing. It starts the concept that IT is a factory and suggests looking at service delivery as a supply chain. It takes the SCOR (supply chain operational reference) model as a basis and uses it to manage and benchmark IT operations. Hans van Aken describes this in his 2011 book The IT Factory.
Fundamentally, the two models describe a series of processes and allows you to measure the performance of each service and optimise it. What tdo we look for in information technology to improve operations and reduce costs? Let us start with IT operations.
Optimising IT operations
I remember a customer where it took 6 weeks to provision a server, when the server was available in the warehouse. We started looking at the steps in the process and came up with following:
- Bring the server to the datacenter
- Get the server in the datacenter
- Mount the server in a rack
- Wire the server
- Configure the server in the network
- Install the operating system, the middleware and the monitoring tools on the server
- Configure the parameters
- Test and QA
- Take in production
In a traditional lean approach, I spent an afternoon asking why, why, why? A couple of things became clear very quickly.
Firstly, the different parts of the organisation were not synchronised to work together and 99% of the time the server was waiting for the next step.
Secondly, there was no end-to-end planning and execution. It was a number of individual work orders that were generated every time the previous step had been performed.
This is just an example, but taking an end-to-end look, basing yourself on described processes and associated measures, allows you bring the time down to a couple days. Think about the improved services and the savings.
Starting from clearly defined processes also brings standardisation with it. The same request is handled in the same way. The organisation knows what needs to be done and becomes consistent in its execution. This reduces mistakes and eliminates the cost associated with it.
If the same process is executed regularly, another option offers itself to you, and that is automation. Why not automate many of the steps? This reduces the manual intervention, reducing cost along the way. It also makes operators more efficient, which in turn reduces cost even more. Operators can now focus on the exceptions rather than on the routine activities. Enterprises having implemented automation have seen a reduction in the number of operators by 50 to 100%.
Measuring the processes, gathering the information and having a scorecard that makes the results visible is a nice way to manage the continuous improvement that ITIL describes. The objective of analysing and monitoring it, is to free up funding so it can be used to create the new services the business is looking for or to transform the environment for increased agility and responsiveness.
Now funding becomes available, we can start developing the functionality of services. The traditional waterfall approach will not work. We do not have six to 12 months to deliver a full set of services. The question here again is how can we improve the efficiency and effectiveness of our developers?
There are two things we can do:
Firstly, facilitate the job of the developers so they can focus on their key task, developing applications. A number of years ago when we analyzed what our developers were doing, we realised they spent around 15% of their time managing the environments they needed for the different steps in their projects. They kept reconfiguring their servers, many of which happened to be under their desks. We concluded that they often used the versions of software and middleware closest to hand, resulting in many versions of the same middleware being managed in production. We ended up providing them with a cloud based development and test environment allowing them to provision environments in a matter of minutes, resulting in more time spent to develop software and standardising tools and middleware. The benefits included time, license costs, but also operation costs once the software was in production.
We complemented this with test automation. Rather than developers spending one day per week integrating the software modules and running test scripts, we automated the process. Developers would clock in their developments every evening, assembling the software modules, a series of tests run and reports generated. By the next morning, the developer would know what was working and what was not.
Not only could he now code five days a week, but also he would not go down a wrong path for a week. Here again, looking at the processes, understanding what was happening, identifying waste and addressing it, helped us free up time and in development this results in speed of delivery and ultimately money. Complemented with agile development methodologies, we were able to address the needs of the business much faster.
A lean approach brings you typically three things, first a deep understanding of how things work and a measurement system to track improvements, second a standardisation of what you do so you become more repeatable and predictable and third the identification of areas for automation. To quote Bill Hewlett, “What cannot be measured cannot be managed”. To free up the funds required to develop what the business wants, we need to improve the management of IT operations and development. Lean is a good way to do so.