OTRS : Five Building Blocks for Successful AI Adoption

Artificial intelligence (AI) is no longer a vague concept of the future. AI adoption has made the leap from a silly theory on which sci-fi movies, like Tron and Space Odyssey, were once based to a practical component of efficient, customer-friendly service delivery.

But, while chatbots and optimized search functions are exciting to customer service organizations everywhere, companies need to lay some groundwork before the benefits of AI can really take root. From understanding how your service organization functions today, to evaluating how success will be measured tomorrow, I think customer service leadership teams have their work cut out for them as they dive into this machine-led world!

From my point of view, these five areas must be considered in order for any company to successfully implement an AI system.

1. Increase emphasis on documentation

The idea behind AI is that the computer learns and gets better at offering solutions the more frequently it is used. Now, the machine is not gathering this information on its own. In order for the machine to learn patterns and determine the proper reply to an inquiry, you'll need to provide it with details about past experiences your agents have had. You do this by sharing ticket data or call recordings with the AI tool (well, more likely with its developers).

What does this mean for your business today? If you're not actively capturing cases and documenting solutions, you'll need to start your AI efforts by investigating ticketing systems, call recording options, interactive voice response (IVR) systems and maybe a customer relationship management (CRM) platform. Begin by gathering the data needed to build the 'intelligence' your business needs to run successfully.

2. Focus on templates and processes

Customer interactions are one type of data: your templates and processes are another. Just like a computer doesn't know what words to use without examining your data, it also doesn't know the correct answer to an inquiry or the right solution unless you tell it.

Step back and look at the how your agents reply today. Are their answers consistent? Are written responses standardized? Does everyone follow the same procedure for addressing customer inquiries? If not, it's time to focus on your internal workflows and build out the templates and processes to support these.

3. Think beyond chatbots

The goal of artificial intelligence should be to enhance the customer experience. That should be the primary factor for using it. So, before you dive into building out a chatbot or investing in sentiment analysis systems, stop and think about the value to the customer.

I was at the SMART 2018 Customer Service conference in April and one of the keynotes was done by two members of Amazon's AWS team who relayed a story about the potential of integrated AI solutions. They talked about pairing machine learning with an airline's IVR and CRM systems. When a customer called in, they weren't asked to 'press one' for help with their flight: instead, they heard, 'We see your flight to Chicago has been delayed. We can rebook you on a flight leaving in 25 minutes. Would you like to change your reservation?' Within seconds, a frustrated passenger was on their way home, and the crisis was averted!

So before you jump on today's latest and greatest AI option, look at the systems you have in place. Imagine how they can be integrated to provide a cutting-edge customer experience. Are there points in your operations where agents are frequently referencing multiple systems? Consider how these can be built into your AI solution so that you can deliver the type of service referenced in the Amazon presentation.

4. Plan for the handoff to live agents

As you think about system integration, also consider the handoff from your AI solution to your human agents. The idea is to improve on your service delivery, so be sure that customers only need to provide their information one time. Account for a seamless hand-off to your live agents when it's needed.

5. Invest in your service team

As machines start to handle routine inquiries, the role of your customer service agents will start to shift. The percentage of more complex cases will increase, requiring deeper problem-solving skills. Agents may be in a better position to help the sales team with upselling and finding new opportunities.

Keep in mind that this technical solution will have personal implications, and think about the issues related to this. Ask questions such as:

• Will additional training be needed to help build the skills of current agents?
• What change management efforts will be needed? (After all, agents may worry about the security of their jobs as an AI solution is presented.)
• Do you need to reexamine your hiring criteria so that new agents come in at a more experienced level?

Before you dive in, take a look at the ways AI will impact the people in your organization - and examine related costs?

Additionally, be forward-thinking in terms of how this shift in inquiry content will impact the way you've been measuring the performance of your customer service organization and agents. Chances are that if your service team increasingly handles more complex cases and becomes responsible for providing even more value to the organization, you may need to question age-old metrics, like handle time. Anticipate and talk about changes that might occur as your AI program hits its stride.

Like any new change - technological or otherwise - there's a lot of preparation required to make the most of it. I expect that there are a thousand other considerations that businesses will go through over the years as AI becomes more and more prevalent in service organizations. But for today, if we focus on documentation and processes, we'll be well on our way to using what's available to us now and realizing the possibilities of tomorrow.