/MHCI Capstone Project /UX Research /UX Design
At the beginning of 2017, Institutional Asset Management (IAM) of PNC, the 7th largest US bank by total assets, came to us for “reinventing” their online service by putting clients in the center. Over the course of seven months, our team, as the first UX team working with PNC IAM, designed CUBE, which supports both self-serve needs with easy-to-use user interface and context-based collaboration between clients and advisors with embedded email. As the first stone for future renovation, CUBE allows PNC’s digital offerings for institutional clients to evolve and adapt to a changing future.
Date 7 months, Jan. 2017 - Aug. 2017
Client PNC Financial Services Group
Team Eunsol Byun, Rob Feeney, Anuraag Jain, Sharon Monisha Rajkumar
Instructor Francine Gemperle, Nancy Duan
In order to bring PNC’s currently fragmented institutional online service together and create a consistent experience, we designed CUBE, which unifies the user experience of four most common banking and investment workflows to fulfil self-serve needs. Meanwhile, to facilitate collaboration between clients and advisors from PNC side, we embedded email into CUBE and adopted drag and drop as the primary interaction to support context sharing, which is seen as the biggest innovation of our project.
From research, we know that institutional clients coming to the platform actually have different goals and tasks. To catering various needs, we designed the landing page to be customisable based on their usage. Information discoverability and accessibility is highly improved for infrequent users.
Our research shows that email is the primary tool used by clients to communicate with their advisors. To make the communication more smooth and information more secure, we embedded the email system into CUBE. CUBE mail extracts only the emails that are related with PNC from the client’s inbox and show in the browser.
With CUBE mail, now context sharing becomes easy. By having the content that clients need to communicate on the left and email on the right, communication of complicated transaction or document information is simplified to a drag and drop action. Instead of referring back to the account info and page number, clients now can simply drag the data into CUBE mail. Long email chains is removed and communication error is reduced.
To solve the problem of no tracking for offline documents, we digitalized some paper work like filling forms and designed progress bar for users to track any changes. Transparency and trust are improved by giving users more sense of control to their own tasks.
Data in CUBE is all digitalized and interactive. Unlike the static data from the current statements, reports in CUBE are dynamic and elements can be directly grabbed, moved around, be deleted right away from the statement, and of course, shared in CUBE mail. Interaction with elements becomes more flexible and is under users’ control.
As the consolidated service platform for institutional clients, CUBE can be integrated with PNC’s internal systems. Clients can get real-time updated data on their documents or any task progress. CUBE can also be integrated with client’s internal system. Clients now can assign approvers with full knowledge of whether the approver is available or not.
In our design, CUBE would become the primary client-facing platform for PNC’s institutional clients with all its core features. But before achieving this, data and access consolidation at the back-end must be done. We made a 5-year roadmap for PNC to implement the whole set of solution. Within one year, the current fragmented legacy systems need to be consolidated to provide unified access, like single sign-on (SSO). Within two years, the common banking and investment workflows can be built up to create a unified user experience for clients’ self-serve needs. CUBE framework can be developed to support such service and collect and learn from the collected client data. In a four to five years timeframe, with data consolidation and machine learning, CUBE would become powerful enough to provide more smart suggestion and more personalized service. More important, CUBE would evolve to become a platform that can provide a interactive hybrid service and adapt to new needs of institutional clients.
At the starting point, we didn’t have a clear mission or specific goals for “reinventing” the online service for PNC’s institutional clients. Without pre-existing insights, we started by walking through the current digital offering and talking with employees at PNC to understand its current landscape.
PNC now has two main business facing institutional clients - Corporate & Institutional Banking (C&IB) for banking, and Institutional Asset Management (IAM) for Investment. Each of them has its own digital platform: the one for C&IB is called PINACLE; and the one for IAM is called I-Hub.
Through interview with PNC employees, we were able to map out the current service landscape for institutional clients at the bank. At C&IB, clients self-serve on PINACLE for banking tasks, while at IAM, clients self-serve on I-Hub but are also equipped with two advisors to facilitate asset management. Also, due to rapid acquisition in recent years, there’re a lot of legacy support systems at IAM. Three high-level challenges were identified at this stage.
1. Fragmented service experience due to multiple support systems;
2. Modern product with legacy scaffolding;
3. Scalability to enable rapid evolution.
Based on this, we established our project mission: To create an asset management experience that can evolve with the changing future of PNC’s client needs.
Before jumping into the intensive contextual inquiry, we first looked into the industry trend to figure out where the industry is heading to and how it would impact PNC’s business. From a time perspective, we found that the standard adoption of client-facing technology of financial industry can be roughly divided into three stages.
Before 2015: a user-centered digital platform, on which clients can complete basic tasks, like PINACLE and I-Hub, but it has already become a basic requirement.
2015 - Now: automation. Driven by machine learning and artificial intelligence, automation is becoming a new industry standard for that it improves operational efficiency and reduces cost for big financial institutions. And data consolidation has become a primary tech investment across the industry.
Now - 2020: hybrid service. But moving forward, what believed would become the new industry norm is a hybrid service that combines the best of automation and human service. Digital platform that can provide greater human access is the new difference maker.
So looking into the future, we learnt that a user-centric digital platform alone cannot make PNC stand out from its competitors. A new paradigm that utilizes the strength of both automation and human needs to be investigated.
Our estimates suggest more than half of the activities people are paid to perform can be automated by adopting advanced robotics and AI — either now, or surprisingly soon. By 2020, we expect that the ‘new normal’ operating model will be customer- and context-centred.
— PwC Financial Services Technology 2020 and Beyond
After getting to know the big pictures, we were ready to start our primary research. We talked with PNC’s institutional clients who use I-Hub and PINACLE to learn about their needs; we shadowed PNC’s advisors and ask them to fill an activity log to to learn about how they interact with their clients; we also got a client survey result from PNC to help identify the big pain points.
With all the qualitative and quantitative data in hand, we started the synthesis process that lasts one week by using affinity diagramming. The bottom-up method helped us identify behaviour patterns and uncover 5 deep insights of the topic.
Having a team of several specialized advisors for each client may cause confusion as the client defaults to contacting a single person.
As FAs are always accessible, clients might treat FAs as their own employees, causing the FAs to spend time on work that does not create value for PNC.
Paper and manual processes are currently painful as there is no tracking or feedback on status of tasks conducted offline.
I-Hub was built for IAM clients but its functionality is more suited for the client's external consultants and PNC employees
A web portal does not support the right interaction paradigm for high value asset management workflows that require infrequent but cyclical use.
In order to generate actionable statements, we further clustered the insights into three big themes.
Since the project mainly focus on the client-facing service, among the three defined problem spaces, we decided to focus on the first two layers, which are poor information discoverability for clients and fragmented information accessibility and sharing for advisors. We then identified two HMW (How Might We) questions for ideation in the next step:
We presented our consolidated research to clients and hosted a visioning session with multiple stakeholders from PNC to envision solutions and asked them to vote for their favourite ideas. We also internally brainstormed a bunch of ideas as a team. A non-exhaustive idea list is as below:
How might we make client-facing information much easier to discover and access, especially for infrequent users?
💡A conversational user interface that clients can query questions withouth manually search
💡Customizable user interface to deliver the most needed information at the right time
💡An expert user interface for advisors to enable them quickly deliver information with keywords and shortcuts
💡A powerful search function to support search of rich information and workflows
💡Embedded function that supports easy access to advisors and easy sharing of problem context for clients
How might we unify information accessibility for advisors and make informaiton sharing easier and more accurate with one-time communication?
💡Modularize advisors’ workflow to enable them track progress of each task and share documents with clients
💡A dashboard for advisors to visualize their workflow so that they better know their work status
💡Smart email to automatically parse related information for advisors
💡Embedded function that supports easy access to advisors and easy sharing of problem context for clients
For all the visioning ideas, we drew storyboards to validate them with both advisors and clients. We tested the ideas with people that we’ve interviewed and also with new advisors who didn’t participate before. From the test result and PNC’s desire, we decided to focus on the client side.
We combined what people like for each idea and finally arrived our design idea: A renovated user-centric digital platform with adaptive user interface catering for different groups of users and allows context-based information sharing.
Before starting granular-level interaction and visual design, our team established five design principles as guidance through the whole prototyping phase.
We followed an iterative design approach with 3 major design cycles. During each 2 week cycle, we designed a new prototype of increasing fidelity and evaluated it with in-person usability testing.
To start with, we made 6 layout versions for one workflow and tested all of them with stakeholders. The parallel design process helped us to compare the sitemaps and visual hierarchies. We consolidated the feedback and redesigned the layout combining the most liked features.
To further test the nagivation, layout, and content, we created wireframe and made it interactive by using Invision. The clickable prototype made the usability testers better understand how the flow works since it helped them navigate with ease.
Finally, to test the visual elements and real content, we created hi-fi interactive prototype with Principle and enables drag and drop interaction before spending time writing real code. This round of testing helped us further perfect the placement and appearance of elements.
We extracted 4 most common workflows for institutional clients from our research, and overhauled the user interface for each of them. Also, we changed the nomenclature of the navigation bar to make it more accessible. But due to the time limitation, this was not an exhaustive website redesign that includes all detail-level changes. It was more about establishing a new user interface language for the platform.
From our research, we know that users who use this platform have different purposes and tasks. Also, some users only use the website infrequently and therefore use it as if they’re first-time users every time. So making information and workflow super easy to discover and follow is very important. In our design, instead of having a one-size-fits-all landing page, we present users their most common workflows based on the collected data and machine learning analysis. Users only see tasks that are most relevant to them and can therefore complete tasks without much effort.
Our research also shows that routing client’s questions to the correct advisor at PNC is a problem currently. Clients are habituated to always reach out to the same person for all their problems, which leads to advisors spend extra time in communication, and their work efficiency is thus undermined. To resolve this problem, the embedded email system could automatically detect client’s questions based on which page they’re on and then route the email to the right advisors accordingly.
CUBE mail, which embedded in the digital platform to facilitate collaboration between clients and advisors, is my personal favourite and proudest innovation among all our design solutions. From research, we learnt that outlook is the primary medium for both clients and advisors to communicate with each other, but it’s not that easy to send complicated information in rich format and therefore induces long email chains to solve a single problem. In our design, by integrating email system into the digital platform and having the context and the communication tool side by side, context sharing now becomes easy.
To support rich information communication, we decided to use drag and drop as the primary interaction paradigm. We refrained ourselves from using it everywhere. The drag and drop interaction is only used on items that carry rich information, like account information, approver information, statement data, etc. Now, if clients have questions about their statement and need explaination, they can simply drag that piece of data to the email box without spending effort to manually type account number and page number for reference. Drag and drop interaction reduces human errors and improve the work efficiency for both clients and their advisors.
We simplified daily dressing pick-up for young female and help them gain insights on their wardrobe.
We designed a mobile app to connect property owners and people looking for parking spots.