AIOps: low risk, high reward. But it’s no magic bullet.

There are upsides and downsides to IT automation. On the one hand, it promises efficiency gains, cost savings, reduced risk of errors and enhanced user satisfaction. The flipside of the coin is that without discipline and commitment it can make your situation worse rather than better.

The first step should always be to put together a game plan to attack either the low-hanging fruit – identifying the manual processes to automate first, getting approval, and pushing delivery – or the processes with the potential for the highest business impact. The upside with the latter approach is that it may have better mid to long-term business case justification and high returns; the downside is that it might involve a long and complex implementation cycle. Whichever approach you choose, the good news is that the risk of failure in AIOps is minimal because it’s passed on to the partner or vendor.

The AIOps maturity spectrum

Businesses can be roughly divided into three categories when it comes to their maturity:

• First-generation outsourcers, typically small or medium-sized enterprises, are risk-averse non-tech-savvy beginners. They have no experience of outsourcing and they lack maturity and automation know-how.
  
  
• Second-generation outsourcers already have some AI capabilities and understand the basics, but don’t have a full stack of AIOps.
  
  
• Advanced, third-generation outsourcers are veterans with perhaps 10–15 years of AI experience under their belts. While they might want to implement AI more widely, they’re shackled by the tug of war between business and IT priorities.

Myth-busting in the new-normal world

The COVID-19 pandemic has busted many a myth, including that service delivery only really succeeds when people are working in an office; home workers have proved themselves more than able to do their jobs properly without major disruption to services.

AIOps is not a tool, but a holistic approach to building automation that requires businesses to put in the hard work and develop solid use cases. Businesses want to see the real impact of automation on both their support function and their business operations ­– and they want to see real use cases.

And when businesses break silos – for example, by implementing full-stack automation aligned with their business KPIs rather than focusing on a specific workflow or process ­– they can enhance the positive impact of automation initiatives even further. With AIOps the real trick is to take baby steps, implementing capabilities gradually based on requirements before taking things further to improve other processes.

AIOps is not a magic bullet

Let’s bust another myth held true by many professionals: that AI and automation will magically solve all their problems. Without careful implementation and solid use cases supported by a detailed rollout plan, businesses will quite simply waste time, money, and resources.

The first step should always be to align the capabilities of AI with your business requirements. Take things at your own pace and increase your maturity gradually; look for the low-hanging fruit and automate other processes as you move forward and become more confident and capable. With process improvement, don’t try to do everything in-house if you don’t have the capabilities. Outsourcing is not a sin – done right, it improves your chances of success and shifts the risk of failure onto the service provider.

Recommended further reading: 

Blog: Boost IT operations with artificial intelligence – this is what AIOps is all about

Lesson Learned From an Interview: AIOps Is Not Just About Tools
Research By: Mahmoud Ramin, Info-Tech Research Group, October 26, 2020