If you are a sailor, there is probably nothing more enticing than taking your boat out to a far-off, exotic port.
Picture the scene. Your boat is a beauty – a 70-foot, two-mast Oyster yacht with all the latest electronics. Everything is automated. You take it out on the open seas and cruise into port. That is when things start to go wrong. The port is not equipped with any electronics to speak of. It handles all communications using a paper-based system. So you cannot exchange data. Your stay in the port is denied.
For companies looking to drive value through their cloud operations, this scenario is all too familiar. They spend infinite amounts of time, energy and money automating technologies on one end of the operation, but fail to automate processes down the line, to ensure they are getting the right amount of value out of their cloud investments.
These organizations figuratively own the ships and the ports. Their ships clip along unencumbered – until they reach un-automated, inhospitable ports.
Needless to say, this kind of misalignment can have significant ramifications for an enterprise conducting a cloud transformation. The company is trying to get feature-rich services out into the market faster, because customers are demanding them. But the organization may not be mature enough to withstand a release once a week if it is used to releasing once a quarter or once a year. Technologists get frustrated because their projects get held up. Others in the organization get angry because they do not want to automate tasks and processes they are comfortable handling themselves.
Moving Beyond Manual Tasks
For years, change management functions were all done manually. Network port approvals, data sovereignty requests, data management processes, approval cycles for governance – they were all handled by people. Many still are. If a task is mission-critical, or prone to mishandling, then there should be a high level of oversight. But there doesn’t necessarily have to be.
Technologies exist to set up automated processes that can approve, route, block and/or manage tasks throughout an organization. Companies can create workflows that handle all kinds of procedures in standard scenarios but then escalate transactions to human intervention in certain circumstances.
If the bulk of the menial tasks are handled automatically, organizations can rev their engines and churn out feature releases at a rapid pace.
Imagine if a continuous integration/continuous delivery process were bogged down by manual processes — it would not be a true CI/CD process. An organization has a well-laid-out plan, a top-notch toolchain and professionals committed to writing, integrating and approving code. But leaders in the organization are clinging to a patching process that requires that approvals be handled through ticketing and emails. Because releases have to wait for patch approvals, the process grinds to a halt.
It is a People Problem
Usually, people create the obstacles. As a general rule, people do not like change. If someone has the job of patching servers, they want to keep their job and do it the way they have done it for years. They do not want to cede control to a toolset that automatically approves server patches.
Unfortunately, in certain cases, automation is inevitable, and it is disrupting people’s jobs. IT budgets tend to be composed of 35-40 percent hardware (servers, networking systems, storage, etc.), with the rest allocated to people. Procurement leaders are under constant pressure to drive costs out of the hardware side. All that is left is to look for savings in the rest of the budget through automating human tasks.
Looking again at server patching, there are plenty of ways to reallocate funds in this area. Rather than patching servers regularly, an organization can build “hydration” into the process – essentially draining old servers and rehydrating with new patch servers every 30 days or so. Or they can just leverage serverless technology. This replaces processes with the outsourced automation of a platform-as-a-service (PaaS) program. In other words, someone else will be managing the process.
Sometimes executives are too wedded to the old processes. In these situations, leaders have to act to figure out a way to usher in new sets of automated procedures to move the organization forward.
There are ways to deftly solve situations where executives are blocking automation in their own individual ports. Leaders can swap roles: take a middle manager and give them a different group to run. They can shuffle groups around, blending cloud-focused teams with old-line teams, and in the process, force the change in the way the whole extended group operates. They can retrain executives and encourage them to get on board with the new procedures. Of course, if a light touch does not work, a heavier touch might be required.
To generate value from the cloud, organizations do not need to create whole new processes. Classic IT change management functions are not going to go away entirely. In many cases, auditors and regulators require that procedures be locked down and tracked. What we will see more and more is the need for process automation, not wholescale process re-engineering.
The one thing organizations cannot do is underestimate the effort process automation requires. They need to commit to it and institutionalize it, in order to wring the value out of the cloud that they are expecting.