Omnichannel Customer Experience: Expert Systems, 360 Degree Views And AI

“Omnichannel customer experience” can simply be shorthand for your customer’s ability to order from you in one channel (for example online), pick up through another (for example at a brick and mortar store) and if necessary return via yet another (for example from your porch or a dropoff point).  (Tip: Try visualizing this aspect of the omnichannel customer experience by using my paradigmatic “Megan Millennial” as your omnichannel customer model.)

But an omnichannel customer experience also stands for the ability of your customer to be in contact with your company via a variety of channels and to essentially be able to pick up where they left off on one channel and continue the conversation on another.

There are two basic challenges when input is coming in via myriad channels:  First and foremost, companies have a responsibility (and a business need) to make sure to respond: to respond quickly, helpfully and humanly to input from customers, regardless of the channel it comes in on.  But a secondary business need, one that is necessary to achieve true, sustainable omnichannel communication as part of an omnichannel customer experience is to learn how these channels come together to create the customer experience, and how a more seamless communication landscape can be created.

Pictured: Mazin Gilbert and members of his team at AT&T Labs review incoming omnichannel customer experience analytics

Your Customer Is The Star: An eBook From Forbes How to make Millennials, Boomers and everyone in between fall in love with your business. By Micah Solomon.

(To give a couple very simple examples: Are customers resorting to Twitter because they didn’t get resolution on the phone? Are they calling because nobody answered their tweets? Or do customers simply like to use a variety of channels depending on the situation?)

In light of this, I spent some time interviewing a company whose omnichannel communication challenges almost certainly eclipse yours: AT&T, a corporate giant that is contacted (and in turn contacts) million upon million of customers through multiple channels 24 hours a day. AT&T’s work on a 360 degree view of the customer communication experience fits both with their position as the operator of one of the largest contact centers in the U.S. and also with its stewardship of AT&T Labs (a descendant of the historic Bell Labs), whose assistant VP of inventive science, Mazin Gilbert, filled me in on the current state of their art.

Micah Solomon: Can you talk with me about the analysis you’re doing to improve the omnichannel experience AT&T customers receive?

Mazin Gilbert, Assistant Vice President of Inventive Science at AT&T Labs: We’re among the largest contact centers in the U.S., so we get a significant amount of interaction data from communications with customers and support teams. It’s a tremendous amount of data.

At AT&T, omnichannel analytics is an effort to enhance the customer experience by tracking and analyzing their journey across all touch points with us. That includes voice calls, chat support, web inquiries and social media.

Now that customers have many ways to connect with AT&T, we need to understand how these channels come together to create the customer experience. We have to look at all the customer touch points together – not apart. Customers appreciate having a connected experience. It increases customer loyalty. It’s also important to drive effortless customer care by identifying and removing inefficiencies in our systems.

This is very different from the old approach of customer data coming in a week later or a month later. We can now capture and process the data in near real time. We crunch large amounts of data to get to decisions in a matter of minutes, and ultimately even seconds.

We can bring many different data sources together. Think of it like a jigsaw puzzle. We’re putting all the data together, anonymize them to strip out personal information, and developing an omni-channel view to experience the journey of our customers.

We’ve created an omnichannel analytics web platform and made it available to supervisors, managers and key stakeholders in a secure setting. They can interact with the application and know the pulse of what’s going on in their business and how to improve it. This is very different than the traditional approach where research and technology teams find something interesting, and then they go and share with the business, “Look what we found.” That’s not what we’re doing. We’ve taken everything we just talked about and wrapped it into an application and made it available to the business units, stakeholders and supervisors so they can interact with the data. Basically, we’ve streamlined the end-to-end process to empower all stakeholders in the company to make smart decisions on the go. The data allows them to answer questions they couldn’t have answered easily or quickly in the past [and ultimately] creates a better experience for our customers by helping them to solve their problems quickly and the first time.

Solomon: Does your 360-degree view assist in correcting glitches that would otherwise hamper customer service and the customer experience?

Gilbert:  Throughout the course of normal customer interactions, there are times when our social media teams discover outdated and/or conflicting information on the AT&T website and various support articles. In most cases, we are able to make changes and updates quickly, either the same or next day. In one case, the team discovered a missing field on an online form that was pertinent to customer contact; the field was added within 4 days. In addition, a process for customers who filled out the form prior to the correction was put in place. There are many times when there is a process needed that does not yet exist.

Once, during a major device launch, we learned the tax calculator in the shopping cart wasn’t working. We were able to correct it immediately – within 30 minutes of being notified. We are always looking for ways to make our website experience better. If we see people having trouble with a link or an order, we notify the correct process owner to have the root cause fixed as soon as possible.

Omnichannel analytics will help these social teams improve the current approach to online customer care. It will help them identify and solve issues much faster, because they’ll have all the analytics they need right at their fingertips.

Solomon: Anything else that’s new and wild that you’re cooking up to improve the customer experience?

Gilbert: Omnichannel analytics is [just] one part of the company’s overall drive towards a digital experience, meaning AT&T is streamlining services to provide effortless and compelling customer care. It’s also important as we transform our network into a software system to enable a rich and a large set of applications and services. The focus of the department is driving autonomous services through technologies such as machine learning, artificial intelligence, recommendation and on-line learning. In artificial intelligence, an expert system can learn from customer interactions just as a human expert can. The systems can imitate the decision-making ability of a human to provide proactive, actionable responses to each customer.

These ‘expert systems’ are continuously learning from human data to provide smarter recommendations. Although we’re driving autonomous services in our contact centers, not everything will be autonomous. We’re capturing intelligence so we can do our jobs better. It allows us to give our customer care agents better information on the spot, and it keeps supervisors and managers informed about what’s going on.

Solomon: Now you definitely have my attention: Tell me about AT&T’s work in Artificial Intelligence (but please don’t tell me anything that would require you to kill me!)

 Gilbert: We’re using artificial intelligence and machine learning to allow us to predict events before they happen and act more proactively to best serve our customers. So for chat, we’re predicting if we resolved the customer problem or if the customer is going to call us again within 24 hours. Our goal is to help resolve the problem the first time the customer contacts us. We’re using artificial intelligence and machine learning to help us predict the emotional state of the customer, because unhappy customers leave us. It’s easy for someone to look at a chat and say, “Yes, we addressed the problem.” But doing that at scale and in near real-time is impossible for any human. With this new technology and a big data approach to customer problems, we can process data quickly. From there, we can make accurate predictions. It allows us to not only identify what happened, but to know why it happened and what the best approach is to solve it the first time.