AI gives IT an edge in improving customer experience

Even prior to the pandemic, improving customer experience was becoming a major priority for IT. COVID-19 and the resulting shift in business models have only accelerated that strategic directive.

Delivering a positive customer experience is even more important now, as businesses prepare for a post-pandemic world that will still involve lots of home-based workers, rising e-commerce transactions, and an unprecedented number of digital interactions between companies and their clients.

From healthcare to retail, artificial intelligence is rising to the challenge. With AI, enterprises from an array of industries are improving customer experience by accelerating how tasks are handled, gaining more insights about customer behavior and preferences, and providing new interfaces to make transactions more pleasurable for customers.

Here is a look at how four companies are deploying AI to deliver a better customer experience.

Virtual waiting rooms and texting chatbots

Banner Health, a healthcare company with a network of 300 clinics and 1,500 physicians that serve more than one million patients across six states, has rolled out “virtual waiting rooms” using technology from LifeLink that provides texting-based chatbots.

Mobile chatbots interact with Banner patients in a conversational style to help them complete digitized forms, provide education, and enable remote check-in capabilities for all telehealth and in-person visits.

The effort is part of Banner Health’s strategy to “re-imagine” how it delivers care to patients in response to the pandemic. Healthcare organizations need to quickly innovate to ensure all patients can see their doctors through safe, private and convenient channels, says Jeff Johnson, vice president of digital business at Banner Health.

The traditional pre-visit process of walking into a waiting room, filling out paper forms, and waiting for an exam room to be available had to change, Johnson says. The chatbots have already helped hundreds of thousands of Banner patients with emergency room visits, so the idea of expanding the technology to regular doctor visits with a mobile chatbot assistant was natural.

The chatbots welcome patients for telehealth virtual and in-person visits with their primary care physicians and specialists, interacting with patients on their mobile devices or desktop computer using automated conversational messaging to remotely complete tasks previously handled in person.

“We know that customers of healthcare expect the same easy-to-use experience that they are getting from service providers in other parts of their lives,” Johnson says. “We need to reduce and eliminate the complexities of accessing and managing care, and our digital strategy supports the types of interactions that can provide that ease — whether it be a mobile app, voice assistant, text message, chatbot, or even a phone call.”

Another area where Banner is seeing a lot of value is with AI-supported digital triaging services. “We know that oftentimes our customers are seeking direction about their health issues or symptoms that are affecting them and where they should go as a first step,” Johnson says.

The company worked with Buoy Health to integrate its AI symptom assessment product. “While many triaging tools emerged during COVID, we actually began our work with Buoy early last year and designed it to be a broadly focused customer support tool to help them in their decision about what type of care they need and where they can receive it,” Johnson says.

When individuals interact with the symptom checker, Banner provides them with possible causes of the symptoms and how they can be best addressed, whether it’s at one of Banner’s urgent care sites, a primary care doctor in its medical group, emergency rooms, or through self-care.

“And the recommendation is specific, so if urgent care is a recommended option we look at the customer’s location and connect them into our urgent care scheduling tool to show the closest location, the next available time, and an option to book the appointment all in one easy digital transaction,” Johnson says. “As we move forward, we will include cost and price information into the care options to provide complete customer decision-making tools.”

Banner does not look at AI as singular strategy, but “it can play a role in potentially all of the digital services and interactions that we provide by allowing them to be hyper-personalized to the customer,” Johnson says.

Analyzing online actions to enhance service

In today’s rapidly transforming retail landscape, turning website visitors into loyal customers is vital. For companies such as Rack Room Shoes, ramping up the e-commerce strategy to drive revenue has become imperative to survival during the pandemic.

The footwear retailer began using AI-based tools prior to the pandemic to gain insights into its customers’ experiences when shopping online. But the health crisis “significantly accelerated consumers’ move to e-commerce, and in turn it’s accelerated our need to understand precisely how our customers are interacting with our e-commerce platforms,” says Kevin McNall, director of digital projects at the company.

Rack Rooms Shoes is using AI technology from vendors such as Dynatrace to automatically analyze every customer journey, down to the impact of every click, tap, or swipe on its website. The platform from Dynatrace provides alerts with the precise root cause of any anomalies or problems that arise, which helps to reduce false positives and enable the retailer to deliver smoother customer experiences.

“We are relatively early in our e-commerce journey,” McNall says. “Yet by continually finding ways to optimize our customer experience [insights from AI] have enabled us to introduce new revenue generating features to our online store much faster, and have improved conversation rates by 25 percent.”

Rack Room Shoes has taken a deliberate approach to customer experience that focuses on the user, McNall says. “All our designs are centered around user behavior,” he says. “We have three key pillars that guide everything our team does: early and continual focus on user actions and their behaviors; empirical measurement of those behaviors; and iterative design.”

AI tools have been integral to supporting these pillars, allowing the company to accurately measure response times and clicks that are then used to inform design. “From there, we iterate, test, and push everything back out to our platform,” McNall says.

Given that “AI” has become somewhat of a buzz term among vendors, any application of technology said to be powered by AI needs to be based on practicality and have real-life use cases, McNall says. “It needs to drive measurable improvement to the customer experience and help us grow our business,” he says.

Making online car purchasing easier

For Cars.com, which operates a digital marketplace for the automotive industry, the customer experience challenges posed by the pandemic reinforced how important the company’s existing AI strategy is, says Mike Hostetler, vice president of enterprise technology and operations.

“We’ve seen in our consumer data that COVID has really changed the expectations of how customers are looking to buy a car,” Hostetler says. “Dealerships are rapidly transforming to serve customers through digital channels. COVID accelerated this transformation.”

At the same time, Cars.com has seen the pandemic move up the purchasing timelines of many car buyers. A lot of consumers who do not own a car are now considering buying one because of decreased or discontinued use of public transportation, and they’re looking to conduct transactions online and relatively quickly.

Once consumers are on the Cars.com site, AI-powered tools help guide them to relevant information based on their needs. One is Conversations, an online chatbot that greets shoppers, answers general questions, and captures lead information on behalf of car dealers so they don’t miss connection opportunities.

The use of AI on the site is designed to make life easier for consumers. “For example, when a shopper starts to look for a new midsize SUV, we make sure that their experience as they go through our website offers them midsize SUV specific information and the brands that carry these body styles,” Hostetler says.

As consumers move through the Cars.com shopping platform, their actions are captured via the company’s proprietary activity tracking tools, which then make recommendations of which vehicles shoppers might be interested in buy.

“The hardest part of car shopping is narrowing down your choices,” Hostetler says. “Our Best Match tool uses an AI-powered algorithm to help car shoppers narrow down their choices based on the features, body style, pricing, color, etc. This allows our shopping platform to only show shoppers the cars with the features that will meet their needs and rank them in the orders of relevance.”

Since the start of the pandemic, consumers’ expectations of a digital marketplace, consumer behavior, and the consumer experience have all changed, Hostetler says. “We’ve started making enhancements to our algorithm to help shoppers find the right vehicle,” he says.

With the growing need for fast and effective service, Cars.com has seen an increase in demand from dealers for products that help them better engage with customers online, Hostetler says. Dealer Inspire, a technology company Cars.com acquired in 2018, has seen a 65 percent increase in inquiries from auto dealers about AI-powered chat tools during the pandemic.

The pandemic has accelerated the digital retailing strategy for many of the dealers Cars.com supports. “Dealers are now stepping up to the plate and optimizing their websites to get consumers more and more comfortable with conducting much of the car buying process online,” Hostetler says.

Moving data entry to bots to add efficiency

As part of its customer experience strategy, Intermountain Healthcare has deployed technology for automating the delivery of acute and ambulatory registration and clinical forms.

Using robotic process automation (RPA), the company sends personalized texts to patients containing the clinical forms needed for specific types of visits, says Ryan Smith, CIO at Intermountain.

“Registration will surface things such as consent or insurance, whatever is needed for a particular patient,” Smith says. “We are reducing the time needed in clinic ahead of the visit, which provides tremendous patient satisfaction. Our patients like completing this work when it’s convenient for them.”

On the system side, Intermountain is moving the task of data entry away from caregivers to bots, which results in significant efficiencies and savings, Smith says.

With its automation efforts, Intermountain focuses on user preferences and simple rules engines. For example, AI determines the best time to text a patient.

“Our CX [customer experience] strategy focuses on how the consumer benefits,” Smith says. “Our goals include more time spent face-to-face with the patient, versus having a computer between the physician and the patient, and also for care to be delivered more safely, reducing potential errors.”

Intermountain is considering other ways it might introduce automation and AI into provider workflow. Provider orders and note automation is one possible area, Smith says. “Over time the bots will learn the clinical triggers and inputs for orders and notes,” which will be entered into the company’s electronic medical records (EMR) system, he says. “And as we teach the bot what was approved, changed, or denied by the provider, we’re able to reduce the error rate and create more tailored approaches to each provider.”

Another area the company is exploring is the use of optical character recognition (OCR), the ability to use visual perception so a bot can infer insights. For example, it might be able to read a PDF to find a pharmaceutical brand name and suggest a generic equivalent.

And like Banner Health, Intermountain has introduced an AI-powered symptom checker, which has allowed it to deliver patients to the right venue of care based on their symptoms. This had allowed the healthcare provider to reduce the call volume of its nursing hotline by 30 percent.

Going forward, Intermountain will apply machine learning to the data sets gathered about diagnostics to inform followup questions about care. “This flow will direct our patients to the most cost-effective point of care and guide them in the journey for scheduling, or messaging, or connecting by video visit,” Smith says. “Treatment will be provided as seamlessly, quickly, and comfortably as possible.”