, Aflac, and Verizon: three different Pega journeys

Pega offers a wide range of products, as reflected in this month’s PegaWorld’s iNspire. These include everything from real-time journeys for customers to back-office automation. We sat with three Pega clients and listened to their different experiences.

We started by talking to the oldest customer of Pega.

Citi and Pega celebrate a ruby anniversary

Promiti Dutta is the head of analytics and technology for Citi’s personal banking division in the United States. She began her Pega journey when she joined Citi four years ago.

The analytics group that I’m a part of is responsible for ensuring data and analytical capabilities are shared across the company. We were aware that our decision engine had reached its end of life and we needed to replace it. The first interaction I had with Pega involved individuals who tried to sell us on the new Customer Decision hub. We did some research, because Pega is not the only company that offers this. Salesforce, Adobe, and other smaller companies also offer similar products.

“So which partner would we like to work with? “So, which partner would we like to work with?” What partner best matched our vision with their capabilities at that time, four years ago? Pega was the clear winner for this.”

Citi has been using Pega for many years, including various workflow tools and Business Case Management. It wasn’t a newcomer to decisioning. At one time, Citi was using Chordiant (the BPM and CRM software that Pega eventually acquired). Dutta said, “We had customer conversations before Pega’s decision engine. Just not as sophisticated as it offers.”

Pega Customer Decision hub uses AI to identify next-best actions for each customer in real time. Citi uses the Hub in a more limited way.

Dutta explained that the decision engine does not decide what we offer our customers. We have developed a number advanced methods to determine what we offer. The Decision Hub is used for determining the “when” and “where”. The ‘what’s’ are all loaded into an offer palette. Using contextual clues and models, the decision engine determines when the customer will see the offer.

Citi has already made predictions about the needs of customers, whether it’s a product, an offer, or another form of engagement. Dutta explained that Pega’s engine makes contextually relevant decisions based on the fact that it knows that you are qualified to receive a product, an offer or some other form of engagement.

Citi, like any other financial institution, is extremely cautious in the way it interacts with its customers. It adheres to strict protocols for model risk management, fair loaning, and privacy. This does impose some limitations on the use AI. “Anything that feeds our Pega decision hub undergoes the exact same scrutiny.” To ensure customers were not adversely affected, we had to run the entire decision engine though that same process.

Dig deeper:


Pega

AI will drive the autonomous enterprise

Verizon: Hyper-personalization for business and consumer

Tommi joined Verizon Business Group before Verizon began its business journey. Michael Cingari (now VP of Marketing Science, CX, and CRM) had been using Pega’s Next-Best-Action solution in the call center for the consumer business several years before.

Marsans, a marketing tech strategist, said: “I was acquired by Verizon through XO Communications.” When Verizon 2.0 reorganized us, Mike Cingari created a marketing science practice and pulled us to there to do a Pega business implementation. This was in 2019. We took a long time to get going, but after we had started and our business case was approved, it only took us 13 months before we began to see a return. We broke even in the first year and then 20X the next year.

Marsans and her Pega team worked in the reactive decisioning area, determining the next-best action in response to the customer’s behavior (in this instance, business customers). When someone called the call center to disconnect, a next best action would be available for them. We began with growth opportunities and upgraded; we then moved into digital and non-assisted spaces, and continued to grow from there.

She was asked to explain how Next-Best-Action impacts customer service. “We’re seeing a difference in assisted channels where service reps are always looking for the best offer to please the customer. They never look at other options. We gave them options and they took them up. It was just as effective. Solving a customer’s problem, instead of just paying to keep them, is a much better user experience.

Marsans stresses that customer decision-making is highly personalized. It’s not about what we’d like to discuss with them; it’s about the next-best offer that we think would prefer. There are more than just offers, especially in the business world. There are also ready-made solutions. “We talk to them about what’s the next best of that.”

For the Customer Decision Hub, to make an informed decision on the next-best actions, it must be trained in what has worked before. Marsans said that if you have a transaction history, “you can feed it and basically jump-start the engine.” We also feed traditional regression models into it. Just recently have we begun to use adaptive modeling (AI in the decision hub). It was us who had to learn, and not the machine. We needed to be able to understand how to present offers, what the best sequence of events is, etc.

You can reuse it, regardless of the business case or use case. You can use it as a base for other things. I don’t believe you need a complete implementation that reaches every channel. “I think you can begin where you are.”

How difficult was it, finally, to convince marketers to adopt a mindset that is in many ways counter-intuitive? Marsans said that the dream of any marketer was to have a customer journey, and to be able influence them along the path to get them where they wanted to be. It’s difficult for them to see it as an ongoing conversation that spans multiple channels instead of “I have to send you something and you need to reply to that.”

Dig deep: Mitigating risks of generative AI through a human being in the loop

Aflac: reducing the time it takes to determine value

Aflac uses Pega in a completely different way than Citi or Verizon. It is just beginning to explore the possibilities of Customer Decision Hub. Pega was deployed primarily to automate and analyze business processes. Pega’s App Studio, which is a low-code application development tool, has been used extensively to develop applications that automate and understand business processes.

It’s an initiative that is aligned to our One Digital Aflac Strategy,” stated U.S. CIO Shelia Andersen. “I would say that the journey took about six to seven years. We focused on ways to automate some of our technical data issues and legacy issues.”

Anderson is relatively new at both Aflac Pega. “I’m learning. “I’m still learning.”

The biggest change I see is in the expectations of the engineering staff. Engineers enjoy writing code. There’s been a shift to make them see the benefit of not having to start from scratch. A lot of the foundational work was already done.

Low code has opened up new opportunities for business users. Aflac ran a recent “Pegathon”, where business users were given the opportunity to use App Studio and create apps that addressed specific use cases. There are more planned. It’s an immersive way to get some of our users used to the tooling and to take advantage of that low-code development approach. They can also see the value of what they are able to create themselves.

Pega’s impact on claims processing is one of its many benefits. Anderson explains, “We discovered that we spent a lot more time on claims with lower complexity (and also lower dollar payouts).” After analyzing the situation, we decided that it would be better to automate payment of these claims. To automate the payment, we use AI or machine-learning and a workflow to automate. This has been a great simplification for customer service representatives, allowing them to focus on complex and critical cases.”

Anderson has created a team that is focused on generative AI. It’s important to ensure the safety of Aflac’s data and monitor its use. She also created a Pega Centre of Excellence and Community of Practice. “That is a big part of where learning has taken place.” In that community, we have individuals who have been with Pega for seven years and others joining the group.

Aflac’s use of Pega, however, was the key to its tangible impact. It allowed it to consolidate several customer service applications across multiple screens onto a single platform, and simplified the work of the customer service representatives. Anderson reports that handling time was reduced by 33% for claims requests, 65% for customer authentication and 77% for all chats. This represents a savings of $4 million.

Anderson spoke on the PegaWorld stage about “shortening time to value and keeping a customer-focused lens.”

The article Citi Aflac Verizon: Three different Pega Journeys first appeared on MarTech.