Natalia Kossobokova is the Content Marketing Manager at CGS. She spearheads the development of global marketing content which includes videos, blog posts, newsletters, editorials, emails and other marketing projects. 

Written by

Natalia Kossobokova
July 30, 2019

How to Build an AI & Human Hybrid Solution for the Best Customer Experience

“With great power comes great responsibility.”

According to Customer Experience expert Adrian Swinscoe, this quote rings true when it comes to the use of AI in customer service centers, just as much as it applies to Peter Parker in the 2002 film Spiderman.

Adrian is an advisor, speaker and best-selling author of How to Wow and Now Punk CX. He has been a long-time Forbes contributor and podcast host who specializes in customer service. Adrian took some time to sit down with CGS to talk about the tips and considerations for implementing AI as part of your customer service strategy.

What are some of the ways that you believe AI could deliver a more robust customer experience?

The possibilities are vast in terms of the potential ways AI can create a more robust customer service experience. With the right access to data, AI can be good at spotting existing problems that have gone unnoticed before. It can be also useful in helping customers predict and find the right answers.

The problems many companies face today is that they think that “If I buy this fancy piece of tech called AI, it’s going to solve all my problems.”

That’s not quite the case.

The first thing companies should do is to build self-help resource centers. These can come in the form of forums, web portals or FAQ webpages. Even though companies say they’re customer-centric, so many of them bypass this step.

Solve problems digitally before they arise or escalate.

Building self-help resource centers can be quite labor-intensive, mundane and time-consuming. In this case, you can use AI and predictive analytics to build in an “auto-fill” option for these self-help centers, allowing for faster search results and quicker problem resolution. See Dell’s support page as an example. To this day, it’s astonishing just how many companies don’t offer a search or self-help option on their websites.

If you produced the right info available to the point where machine learning can help them answer the question, it will make the customers happy. It would alleviate all those repetitive and banal inquiries that come into the call center repeatedly. These are the type of questions that drain resources.

There’s lots of value in an integrated, self-help and AI approach.

The second critical piece is understanding how AI aligns with your commercial strategy. It’s easy to be distracted by the potential of AI, but it’s also important to lay out a detailed plan of how this is going to drive business. How will this increase revenue? How will this allow my company to grow in the years to come? It’s important to think about the end result. Think of how AI fits into your strategy, and not how to fit AI into your business. Otherwise you end up with the tail wagging the dog, so to speak.

When going through digital transformation to incorporate AI into customer support processes, do you have any advice on how companies can maintain focus on the human experience?

You have to really understand what the customer goes through and what the employee goes through. Ask yourself: How can we remove the pain points that both sets of people experience? You must invest time and resources into studying and analyzing the customer support journey. It comes from understanding the human experience.

When you think about proactivity, many companies are still in that “how can I sell you more stuff” mindset instead of “how can I solve this problem for my customer?” Do the research. Create focus groups. Use data from your call centers. Do the work upfront to reap from the benefits later.

How can AI assistants enable human agents?

Having an AI assistant for your agents allows them to find answers to questions quicker, improve first call resolution and increase customer satisfaction. It improves performance and efficiency overall.

One good example is CSAT.ai, which an app that sits within a help desk or email software. When it comes to email, it can adjust your grammar, make you sound more empathetic and empowers learning on the job.

Another example is Cogito, a real-time conversational guidance and analytics tool that listens to calls between agents and customers. It also analyzes the structure of the conversation and provides real-time emotional intelligence coaching on the call. In practice, it listens to the call and suggests prompts like: “Slow down. Let the customer finish talking.”

These types of AI assistants help agents in contact centers with really difficult conversations, removing tension, stress and elevated tones – which is especially helpful for healthcare and financial companies. Imagine a conversation with someone who is elderly and is experiencing healthcare issues speaking to a young adult. The customer calls in and they’re quite fraught. The assistant helps the agent by reminding them to pause, be more sympathetic and suggest phrases along the way. Implementing this kind of AI assistant in contact centers can result in increased customer satisfaction, employee satisfaction, reduced stress levels for the agents and lower employee turnover. You can check out a recent interview on my blog for more about this.

Moreover, further research shows that even though we know being empathetic is a good thing, it’s not something we often do well or default to. The reason being is that our brains are wired to be lazy or default to our natural habits. Our brains are wired to act on instinct because it conserves energy. When faced with a situation whether we should be empathetic or not, it requires effort.

Have you seen any big mistakes in companies implementing AI in customer service?  How can they be avoided?  

Ethics is one of the biggest concerns.

Some of these advanced applications of AI are incredibly powerful. The question is: with all this technology, just because we can – should we?

There’s a continuum between more open systems that are easily understood to more opaque systems which can produce brilliant outcomes and also outcomes we don’t understand. Some of these systems can create some horrific outcomes based on how some of these AI models were produced or the data that they are fed. Understand the risks involved and ethical dimensions.

Another concern is, AI is turning into a “catch-all” term that is becoming meaningless.  Kind of like “social media” was seven years ago. AI is also in danger of becoming this kind of overused term. Right now, it’s more about the technology than what it can do. All of this should never just be about the technology. AI will change. That’s an absolute given. It would be more useful if we focused on the outcomes rather than get completely obsessed by the technology.

How much personalization should customers get during their customer service experience? What should companies do to ensure their clients trust them with their data?

This is where we come to the personalization paradox: Over 75% of customers will say they want a personalized experience. Yet over 50% say they have concerns about data privacy and security. However, one cannot exist without the other.

It’s best to look beyond the stats. Customers may not be thinking about the implications of you having their data. When in doubt, ask your customers questions like: “If we collect XYZ data… would you be OK with ABC personalization?”

Companies need to stop being so scared of customers. If they have any doubts, they should test it with customers so they can get better at self-regulating. Don’t guess. Take the time to ask people. When it comes down to it, it’s humans helping humans after all.

 

Natalia Kossobokova is the Content Marketing Manager at CGS. She spearheads the development of global marketing content which includes videos, blog posts, newsletters, editorials, emails and other marketing projects. 

Written by

Natalia Kossobokova

BPO study reveals biggest challenges for growing tech companies