Today’s digital business strategies come with a list of enticing expectations: improved process efficiency through automation, increased employee productivity, better management of business performance and new revenue streams, to name a few perks. But with all of the promises offered by digital strategy, there is one simple truth:
“The people are the center of any digital transformation. We understand that it has to work for our users or we’re not actually solving their problems or making their lives better,” says Mouneer Odeh, vice president of enterprise analytics and chief data scientist at Thomas Jefferson University and Jefferson Health in the Philadelphia area.
Organizations are finding that to be successful, they must improve the digital experience. Many of this year’s Digital Edge 50 award-winning organizations have done just that. Here they share how they’re using new technologies to enhance how customers experience their companies’ products.
Rescuer network puts motorists in good hands
Digitization has revolutionized the way Allstate delivers roadside assistance to members and nonmembers alike. The insurance company is even sharing its mobile technology with industries outside of its traditional circles.
Anyone who has been stranded in a broken-down car on the side of the road knows that minutes can seem like hours when waiting to be rescued. Allstate has been working to reduce the anxiety and uncertainty that come with roadside assistance since 2015, when it rolled out its pay-per-use Allstate Rescue Mobile App. Any motorists could download the app and, with a couple of clicks, get help from a reliable service provider with pre-negotiated rates and even share the status of their rescue with family or friends.
“We have stripped down and thrown away our old ecosystem that essentially was built to ‘dispatch’ help” in an analog fashion, says Kamal Natarajan, vice president and divisional CIO of Allstate Roadside Services. Today, “we pride ourselves in offering humanized rescue experiences for our customers using digitally enabled solutions.”
While the Rescue Mobile App was successful, it still didn’t resolve the amount of time motorists had to wait for help. Service providers and towing companies typically earn more money responding to municipal 911 calls than roadside calls, Natarajan explains, so they would automatically give stranded motorists a lengthy estimated time to arrival, usually about two hours.
In 2016, Allstate rolled out the Good Hands Rescue Network, the industry’s first crowdsourced rescuer network that can assist stranded motorists by bringing gas, changing a tire, jumping a battery or popping a lock — anything that doesn’t require a tow truck.
“We were able to reduce ETAs,” Natarajan says. “What used to take two hours now takes less than 30 minutes, and it’s a lot cheaper for customers to get that kind of help.”
Today, the network includes about 2,200 active rescuers. The contractors are typically retired professionals, often car mechanics or veterans, who are looking for part-time work. They are background-checked and trained by Allstate before being allowed to respond to service requests through the app.
In 2017, Allstate decided to convert its mobile assistance technology to microservices and make it available to external partners as APIs. It also introduced a developer portal that allows Allstate partners and prospective partners to take the APIs for a test drive.
Not only have the APIs attracted the interest of car companies, ride-sharing services and roadside service providers, but Natarajan was surprised by the interest from other consumer industries, such as family networking and location-sharing apps.
“For these companies that are promoting a sense of community, if they can provide a digital experience that helps manage anything that causes anxiety, then they want to offer it as a value proposition,” Natarajan says. These partners can offer the roadside mobile app as a “vanilla” or a branded pay-per-use offering in their own product.
His advice to other CIOs embarking on similar projects: “Don’t get carried away with thinking of it as a technological idea. Focus on the customer experience, and you will do wonders.”
Improving patient outcomes with data analytics
Jefferson Health is saving lives and improving outcomes by centralizing data and adding an analytics platform. Thomas Jefferson University and Jefferson Health was formed by the merger of five independent health systems and two universities within three years. Its combined data and analytics capabilities were fragmented, inconsistent and not aligned to the organization’s strategic priorities.
The Jeff Insights Platform was created to make it easy to access and analyze data, deliver value quickly, and support innovation through the use of machine learning, artificial intelligence and predictive modeling capabilities. To be successful, the platform needed to deliver the right insight to the right person at the right time in the right context, so the analytics solution was embedded directly into existing workflows, such as through its electronic health records (EHR) system.
“The technology is humanizing not only for the people who are receiving the information and acting on it, but also humanizing and better improving the lives of our patients,” Odeh says. “That is ultimately what we are trying to achieve.”
Shortly after rollout, the platform began showing benefits. For instance, Jefferson Health quickly improved colorectal cancer screening compliance from 60 percent to 67 percent by identifying gaps in care or patients who were approaching gaps in care. The platform helped Jefferson create a registry of all patients with medical conditions that make them candidates for screening. Clinicians would see a reminder in the identified patient’s EHR chart to discuss screening during their visit and make services available immediately. This resulted in 15 fewer diagnoses, five fewer deaths, and $600,000 in reduced costs per year.
The technology also helped Jefferson Health develop an Opioid Provider Scorecard that led to a drop in unnecessary prescribing. The Jeff Insights Platform helps the tech team identify patterns. It identified in the EHR system that the length of prescription had a default setting of 14 days. New medical guidelines recommended doses limited to three days. Within a month of the default value being changed to three days, the emergency room cut in half the number of opioid prescriptions of seven days or more. “This is encouraging us to think of other ways to use analytics to identify opportunities to better humanize the EHR experience for our clinicians and ultimately for our patients,” Odeh says.
The new platform has helped drive $108 million in financial impact so far, but it’s still a major cultural change for the organization, Odeh says. He continues to drive engagement with the platform.
“We’ve stopped thinking of ourselves in terms of delivering content and now think about how we can drive engagement and impact — to change the experience from feeling intimidated by the data to feeling empowered by it,” Odeh says. “Our success is about more than just delivering analytics and checking the box. If we’re not thinking about the people who are using our solutions, we’re not going to be successful.”
Bridging social services with a new service delivery model
Michigan’s governor challenged its state agencies in 2015 to knock down the silos that divided social services providers and deliver a more person-focused experience to its residents, rather than a program-focused delivery model — one that would understand the whole person and what’s holding them back from success.
The MI Bridges digital platform was developed to address these goals. It provides residents with access to more than 30,000 state and local resources such as food banks, shelters and childcare providers. Using MI Bridges, customers can explore resources offered by local community organizations, apply for benefits, report life changes, make case updates, or renew benefits — from a desktop or laptop device. They can access their own benefit information, letters and verifications sent from the Michigan Department of Health and Human Services directly from their profile, without having to contact a caseworker, and they can upload documents directly from mobile devices to share with MDHHS. A pilot version of MI Bridges went live in 2017 with a phased statewide release completed by April 2018.
“We’re hoping to have a fully integrated HHS platform, so someone in need of services is able to ultimately go to one place to get the services that they need,” says Ward Beauchamp, general manager at Michigan’s Department of Technology, Management and Budget. “We are taking incremental steps to get there.”
The application for state benefits used to take an average of 40 minutes to complete. With the launch of MI Bridges, the average application time was cut to just 17 minutes. More than 600,000 residents have registered as customer users of the site, and some 2,220 community partner users and 680 community partner organizations are registered in MI Bridges to partner with MDHHS to service Michigan residents.
“We took the approach of leveraging Salesforce (GovCloud) as a platform for service and an enterprise service bus to share information between legacy systems,” says Judy Odett, business relationship manager. “That allowed us to take an integrated approach. It didn’t require many changes to those legacy systems to deliver that modern view of delivering services.”
The project team relied heavily on regular feedback from stakeholder focus groups in developing the platform. “Sometimes business or IT people think it makes sense, but when you get it out to the population, it doesn’t make sense to them. Taking that extra step helped it be successful,” Odett says. The agile process allowed for rapid improvements based on real-time feedback, resulting in a product that is truly collaborative, she adds.
Beauchamp recommends that organizations pursuing a similar path have established governance or a leader who is looking across the different program areas with an eye toward integration. “Each program area is very focused on delivering excellent quality for what they’re responsible for delivering,” he says. “It may be difficult for them to see how to integrate with other areas without a view that brings it all together.”
Enabling equal access in higher education
Cornell University believes that its 22,000 students should spend more time participating in opportunities outside the classroom and less time finding those opportunities. In the past, students would have to traverse multiple, siloed websites to find information on study abroad, research opportunities or specialized study programs.
“We invest a lot in these opportunities, and we want our students to have equal access to them, not just somebody who happens to hear about it or someone who knows the network already,” says Rebecca Joffrey, IT innovation officer at Cornell.
In February 2018, Cornell launched the experience.cornell.edu website, using a mix of content management and customer relationship management technologies to bring more than 600 opportunities to students’ fingertips in one location.
The site presents suggestions and allows users to compare programs and “favorite” them. Students can refine their search by categories such as subject area, college or school, term and location, and then drill down to see additional details, such as cost information and a summary of deadlines.
Cornell plans to use data gathered from the site to understand exactly who is applying for these opportunities and, more important, who is not.
“In a typical technology system, you can only see the population of people who do take part,” Joffrey says. “But because we’re doing this now in a CRM system, we have data for our entire student population, so we can understand where there are gaps in our ability to provide service, opportunity and access” and deliver opportunities in a more targeted way, she adds.
“Now we have a single place where our programs live,” Joffrey says. “We’re retiring all these link farms and websites and going to a central utility or service. That alignment of business processes has been invaluable. Other websites at the university can pull content from this repository.”
The site has welcomed 26,000 visitors since the launch (41 percent are new visitors and 59 percent are returning visitors) and received 316,000 page views. Students from other universities who take part in Cornell programs, high school students and even the general public can use the site for some programs.
Technologies like these can be very beneficial to all higher education, Joffrey says. “There is a different class of technology out there than what higher ed institutions have used historically. We have relied on a lot of siloed solutions, or a business unit solves their entire business problem for a certain office. We now can use best-in-class tools to provide service.”
Increasing access to city government
CIO Ted Ross foresees a day when residents of Los Angeles will be able to make a verbal request through an Alexa or Google Home virtual assistant device to get a pothole filled or graffiti removed, or to order a new trash can. “You will be able to interact with government through simple conversation in the comfort of your home,” Ross says.
That capability may not be too far away, as the city’s Information Technology Agency (ITA) has been working to increase access to city government. Its latest digitization venture, a chatbot named Chip, the City Hall Internet Personality, has already helped answer questions from city contractors, police candidates and tax-paying businesses.
Chip was first introduced in March 2017 through an integration with the city’s contracting opportunity portal, LA Business Assistance Virtual Network. Contracting with government is never easy, so Chip helps answer a wide variety of questions from how to register for new contracts to what is a federal NAICS code?
Because businesses often use off-hours to search for their next job opportunity, the city actively worked to find a way to be there for businesses when it is convenient for businesses to interact with the city, Ross says.
“In our first 24 hours of launching Chip, we were receiving questions at 2 or 3 a.m.,” he says. “We’ve seen a 60 percent reduction in questions that humans had to answer.”
Chip has been integrated into several other websites and apps, including the LA police department recruiting site, joinlapd.org. Chip answers more than 1,000 questions a month for the candidates navigating the complicated LAPD recruit process.
The bot uses automatic speech recognition (ASR) to convert speech to text and natural language understanding (NLU) to recognize the intent of the caller without requiring the caller to speak in specific phrases, which improve the success rate of self-service interactions. Chip can also detect the sentiment of the user, as well as communicate in more than 50 languages.
“We hooked him into a ‘sentiment AI,’ so we can tell when somebody is getting a little irritated in the responses back,” says Joyce Edson, assistant general manager and deputy CIO at the Los Angeles ITA. “We found that [from listening to] a lot of the responses, that we trained Chip in government-ese and not really simple English — so we monitor those sentiments so that we can then target those particular knowledge articles and make them more English-like.”
“I really like the exercise it requires from my technical staff,” Ross adds. “A chatbot has to be conversational and have strong communications skills, and traditionally technical staff don’t necessarily come with those skills, so it requires us to make our content understandable and digestible, which ultimately makes us a better government.”
Theme park visitors gain express entry with facial recognition
Excited visitors to Universal Studios Florida and Islands of Adventure in Orlando want to spend less time waiting in line and more time riding the attractions. In 2017, Universal Orlando Resort launched the Photo Validation system in the two theme parks, allowing guests with express access passes to skip the regular standby lines at attractions.
Photo Validation leverages innovative facial recognition technology in a way never before used in a theme park to quickly and easily authenticate users of the Universal Express pass.
Typical facial recognition systems require specific lighting parameters and have slow, cumbersome interfaces not conducive for theme parks, which operate in a wide variety of lighting conditions. To develop a system that would meet performance needs, Universal Orlando worked collaboratively with a partner to advance facial recognition technology so that it works in virtually any lighting condition found at attraction queues — from bright Florida sunshine to complete darkness — and is optimized for speed and ease of use, day or night. This created a guest experience that doesn’t feel invasive but is a useful part of the park experience.
The system is multifunctional, allowing for facial recognition and photo referencing, plus bar code, QR code and RFID scanning — all seamlessly, within a few seconds and with secure, encrypted templates, which allow for protection of data.
The system is practical and accommodates the natural behavior of simply looking toward a camera at each access checkpoint, a feature that proved challenging at first. “Sometimes the ‘simplest’ assumptions for guest behavior are the most challenging to realize in actuality,” said Al Callier, vice president of strategic innovation, emerging technology and development, in a statement.
The team thought that guests would find it easy to scan their own ticket bar codes and align themselves to the camera. However, through testing, it was determined that test subjects, and later actual guests, struggled to understand and consistently perform the aspect of scanning their tickets themselves and looking directly at the device. They also observed that families and parties often approached the recognition checkpoint as a group, not individually in a linear fashion. “We ultimately found guest self-scanning of ticket media was not ideal for scanning bar codes,” Callier said. A team member with a scanning device was reincorporated into the process to assist with the transaction, “which is much faster,” and floor markings were also installed to visually prompt guests to stand at the ideal distance facing the camera unit.
Starting with a few attractions to help guests and team members acclimate to the new system, the technology quickly expanded to all top attractions with Express Pass access.
Today, the average processing time per transaction is less than 1 second per scan, which makes processing Express Pass guests faster than ever. Some 58 percent of guests view Photo Validation very positively or positively, and only a handful opt out of the experience.
“Using this technology to reference and validate a diverse and international population of guests in the theme park attraction queue setting had never been done before. Neither had the complex combination and marriage of software, hardware, construction, creativity and theming required,” Callier said.
This article originally appeared in the Spring 2019 digital issue of CIO magazine.