November 15, 2018 | By

AI for IT: Achieving ROI is Easier Than You Think

With all the hype around AI, it may seem like getting started takes a heavy lift. In reality, your team can get quick wins from AI with the right approach and investments. Like the old joke, “how do you eat an elephant”, AI can be tackled piece by piece to deliver the fastest and most impactful ROI for IT. Look for those targeted approaches and small wins, and go from there.

You have read or been told of the numerous ways AI may fit in across many of your IT processes. However, many of these approaches take significant investment – in both time, resources, and dollar – and will take time to show benefit. In fact, – 55% of companies installing AI for support fail to see ROI. Obviously, this is isn’t the best way to validate a technology or approach which is new to your company.

One of the things I feel passionate about, working in a software company, is being nimble and being able to test and learn quickly. This is the approach I would suggest to taking when it comes to standing up AI in your organization. Look for solutions which are turnkey or lighter weight but can impact the KPIs your measuring.

In addition to an “out of the box” solution, for the fastest validation and quickest wins, look for tools which have one of three key AI technologies – Natural Language Understanding, automated routing and escalations, and Robotic Process Automation. Implementing one or more of these technologies can deliver quick ROI and change how your offer support.

Natural Language Understanding (NLU)

Users ask support requests in many ways, not always using the same terms or sentence structures. This is where NLU comes in. At its simplest, NLU is used to understand the intent of the question being ask. By leveraging NLU support questions can be directed to the right self-service or route to the correct support channel regardless of how the user askes the question.

NLU is the core component of many AI solutions – which you shouldn’t need to train. By starting with a tool that already successfully understands intent, you don’t need to “train” the bot to start testing and seeing value.  Once you start testing you can, in many product, start teaching the bot some of your companies particulars and help it get smarter.

Routing and Escalation

Not all tickets are created equal. While some tickets demand advanced IT support or hands-on resolution, many are common tasks that should not be escalated or managed by an agent. AI can be introduced to your systems to manage all routing and escalation, using your own rules to push tickets to the right channels for resolution.

For the most common issues, AI can provide self-service or route support to tier-1 resources. For issues identified as critical, from a senior stakeholder, or customer exposed, routing can go immediately to senior support agents. Your most experienced IT resources focus only on critical problems, while all other tickets are driven toward appropriate resolution.

Robotic Process Automation (RPA)

AI can do more than scan emails for support keywords. With Robotic Process Automation, AI can be designed to completely replicate tedious, time-consuming tasks. Undertaking an RPA initiative can be a big project, so as I mentioned before, focus on a few small tasks to begin. With the right approach to RPA in your IT organization, you can refocus resources on key IT tasks instead of repetitive support.

Within IT, RPA can reproduce the processes you use to coordinate support, route and track tickets, and analyze trends and data. After training them once, RPA processes prevent your IT resources from focusing on repetitive tasks without slowing down your support. In fact, Gartner predicts that 10% of IT hires in 2019 will be solely focused on programming RPA processes.

Small lifts, and quick wins can help make AI an important part of your support process from day one. Learn more about how Prompt ai can make AI-enabled support an (easy) reality.

Discuss / Read Comments

Leave a Reply

Leave a Reply

Explore our other AI Insights or Recent Posts.