9 Steps to a predictable IT environment

1. Determine your level 

The first step is to determine the current level. Nearly every organisation is involved to a greater or lesser extent with ITOA without being aware of it. IT departments that still work reactively utilise certain resources or have set up processes that are an initial step towards ITOA. It is important to realise ITOA is not a replacement approach for the current monitoring and administrative processes, but rather an addition to that which has already been set up. Therefore, analyse what processes and tools are already set up to determine how to move forward.

2. Establish KPI's with the business

ITOA is characterised by a strong focus on organisational. The need of the business and the effect of IT on business processes are the central focus. For that reason, “What does a half-second delay on the system mean to the business?” is an important question. What does it mean for the helpdesk employee if he always has several seconds of waiting time when searching in a CRM system? And what impact does the loading time have on the conversion of a web shop? Therefore, together with the business, determine what KPIs are important and what the limit values are as part of that.

3. Define relevant data sources 

Achieving the KPIs requires insights. What insights are those? And what data sources must be consulted for that? This relates to both data from the IT environment (logs, CPU, loading times, configurations, etc.) and business data (what happens on social media? How much electricity does location X use? What marketing campaign is ongoing?). Continuously monitoring and correlating the data yields information based on which predictions can be made on the behaviour of IT in the short and long term.

4. Select a product 

On a technological level, it is necessary to monitor numerous parameters by means of monitoring software and to collect the output in overarching software that saves and combines the data from individual processes. This software must possess a certain degree of intelligence and machine learning qualities, as a result of which patterns and correlations are made automatically visible. Depending on what data sources are defined as relevant, various products will have to collect this data.

5. Choose the right people 

The success of an ITOA strategy always depends on people. It is important to carefully consider the composition of the team in which both people from IT and the business are given a place. IT has knowledge of the tooling and can make the connections with the defined data sources. The people from the business must recognise the importance of IT and be able to embed the processes in the organisation. This composed team forms the bridge between business and IT and have joint responsibility.

6. Determine dependencies

Separate data is not yet information. It is therefore important to determine what dependencies or events must lead to action. After all, the individual parameters must still be correlated with each other. For this, practical experience and the previously selected software that can establish the connections are required. For example, alarm bells should go off if values exceed a critical limit, but also if certain chains of events occur or if other specific situations occur. In short: when is there a connection or dependency and when is action required?

7. Set up processes

ITOA demands supplemental processes. The business must be involved at an early stage to define relevant data sources, but the follow-up on the insights obtained must be determined in advance. Who looks at the predictive information? Is there a single point of truth that all of the involved parties can look at? Having one central truth, everyone speaks the same language and agreements can be made on the actions to be taken. What process is initiated when a risky change is implemented? Who scales up systems if, based on external sources, it becomes apparent that the load shall increase? A physical Control Room can help with this.

8. Involve the end-user

ITOA revolves around an holistic approach. It is therefore essential to approach the IT landscape based on a chain philosophy. After all, various factors can have an impact on the performance of the landscape. For that reason, it is possible that all of the individual technical domains will function properly, but that from the end-user perspective the performance is not in line with that. Therefore, involve the end-user in the analyses and examine the functioning of applications and systems from a trans-domain perspective. This creates greater insight into the performance and the effect for end-users.

9. Get to work!

It is important to realise that a predictable IT environment is not achieved from one day to the next. The strategy can be implemented in phases, depending on what products and processes are already set up. Getting started with ITOA therefore begins simply with the further expansion of existing monitoring processes. When doing that, look particularly at the connection with the business (how can IT optimally support business processes?) and what data is needed for that. Gradually you will become increasingly mature in ITOA with the goal of a predictable IT environment.