Presented by Bold360 by LogMeIn


It’s time for an end of the year CX debrief! Catch up with this VB Live event for a tour of the year’s customer experience trends, a look at how AI made an impact in 2019, and a perfect line of sight into the biggest challenges and opportunities for CX in 2020 and beyond.

Access on demand for free right here. 


It’s still early days for conversational AI and intelligent assistants, and building a truly intelligent assistant requires a lot of talent and investment, as well as backend work on your tech and data before you get to the frontend experience and a natural language bot. That said — it’s a mandatory to start investing, says Ken Dodelin, VP Conversational AI Products at Capital One.

Companies just starting out actually have it easier now since the technology has been evolving at a very rapid pace, Dodelin says. In the past year alone, there have been a number of advances in natural language understanding. The models continue to get better. The data available to run through those models continues to improve. Things like Google’s BERT model, released in late 2018, helps a broader range of folks make improvements to their NLU models and understand what customers are asking, which of course is the first step to having an intelligent AI.

“We’ve steadily improved our NLU and rolled out with an increasingly better understanding of the customer’s questions,” he says. “Overall I’d say we realized the advancements in NLU across the industry, and we’ve been able to do that to improve the customer experience in our primary digital experiences, our mobile app, and website.”

One of the most important capabilities of Eno, the intelligent assistant that lives on Capital One’s digital experiences, is being able to drill down to discrete units of customer intent, which enables the assistant to answer almost any question that is thrown at it. Not only can Eno now interpret the 2,000 different ways customers inquire about their balance, but as Dodelin explains, can ‘chop up’ an answer into many successive parts.

For example, a customer wants information about their rewards. They may start by asking about their reward balance, but then Eno will anticipate related questions like ‘How do I cash in my rewards?’, ‘When do they expire?,’ etc.

“That level of detail to give the customer an answer, versus just routing them to a lot of FAQs, is a great way to improve the experience,” Dodelin says.

But Eno’s new capabilities are what’s really changing the game, and seriously delighting customers.

“Up to this point we’ve been talking about Eno in a reactive sense, fielding questions from customers, but what we’ve leaned into this year has been making Eno a more proactive assistant,” Dodelin says. “Some of the most positive feedback I’ve seen on any product — I’ve been doing this a long time — has come in response to Eno, without being asked, reaching out and providing customers with targeted, personalized, important information about their account.”

Eno lets customers know if they’ve been double-charged, if a recurring transaction just had an increase, and even asks if it should remind you when a free online trial is about to expire, so you don’t get charged, and the customer response has been extremely positive.

The company has been taking the potential privacy concerns around a proactive assistant very seriously, Dodelin says. They’re extremely transparent about what data they use and how they’re using it, and leave the customer in control. Even things like the fact that Eno is not a human — they make sure that customers understand that in their very first interaction with Eno, so there’s no confusion.

“It’s all about building up that trust with transparency,” he explains. “We obsessively test each new Eno insight, including these proactive things, and try to find that right balance through customer feedback and making sure we’re staying within the boundaries for customers, so the value being delivered is appreciated.”

They’re learning that customers expect the assistant to have access to their transactions, monitoring them, and surfacing interesting and helpful things about them, on a global basis, and it’s been demonstrated to them that that’s within bounds.

“‘Keep going’ is the message we’re getting from customers,” says Dodelin.

And get started is the message he has for companies that have yet to begin exploring the ROI of digital assistants.

“The challenge for folks is that this is a longer journey than launching your average widget,” Dodelin says. “But the risk of not getting started is that when you do decide to go, you have a long road ahead to get something good to the market.”


Don’t miss out! 

Access on demand right here.


You’ll learn about:

  • The future of customer engagement
  • The challenges, opportunities, and trends in universal accessibility
  • How digital technology and AI impacts in-store experiences
  • Finding the happy medium of personalization
  • Why privacy and security is a customer service issue

Speakers:

  • Ken Dodelin, VP Conversational AI Products, Capital One
  • Prince Arora, Director, Product Management, Albertsons Companies
  • Katherine Hird, Senior Product Marketing Manager, Bold360 by LogMeIn