ADMIN PORTAL
Lead Designer
UX / UI / Prototyping / Hand Off
2 Months
Primary Contacts + Objectives
One product owner, one project manager
○ Identifying and building client / user pain points
○ Building of functional requirement documentation from a holistic, UX perspective
One lead developer, two additional
○ Understanding capabilities and restraints based on technology used
○ Running QA on implementation and providing design guidance
Financial institutions, internal product employees
○ Leading client discussions to understand and analyze user frustrations / needs
○ Gathering requirements and improvements through impromptu "testing" sessions
I worked with product owners to analyze and break down Amount's current partner processes, researching the common pain points and potential areas of improvement within our onboarding process.
Working with developers, I helped set up an agile approach to design and handoff that worked in tandem with the tight timeline between initial concept and targeted launch date for our MVP.
Pain Points
Terminology common to internal product employees often became insufficient and confusing to partners, leading to an increased need for follow-ups and circle backs.
Due to the repetitiveness and overwhelming layout of the configuration document, often partners would mistakenly leave certain configurations selected that they did not want. If not caught manually, the build of their product would be incorrect.
Though it increased client/company communications, the general process of the configuration collection stage felt like it expected handholding to be more commonplace than necessary.
Adapting to the time crunch we were under as well as working with limited team capacity, we approached the product from a "dive and conquer" perspective, simultaneously working in tandem to gather and analyze requirements, produce designs, and work towards and MVP build of the product.
We segmented out the configuration collection into three stages, each consisting of several sections of the overall product.
The process was intentionally scrappy, intending for the entire team to be able to pivot and adapt based on updated requirements and partner expectations: placing flexibility at the forefront of our structure.
Tech Stack
Utilized as the primary codebase for the Admin Portal, both due to familiarity from the development side as well as numerous Amount dashboards previously being built on it.
Though being used for the MVP, future iterations of the product were meant to utilize React.
Utilized for increased availability of interactions such as expanding and collapsing components within a configurations module.
Utilized to bolster the UI and more closely match CSS elements according to what was handed off between design and development.
Because the timeline we were working within was both tight and expectant of an agile work process, it was crucial to be able to map out product navigation and effective "clustering" of modules within the portal. I worked with product owners to establish a realistic ordering of modules, and on my own sketched/wireframed out potential layouts for the portal.
Input-related configuration collection through a static and straightforward layout.
Allowing users to rank and prioritize the specific configurations of a module.
"Switching" of configurations on/off in a streamlined manner.
The content of some pages had requirements that exceeded the templates I had defined. In these cases, I set up isolated workflows between myself, product owners, and developers in order to further understand and ideate on potential solutions.
We gathered feedback from both internal users of the current onboarding framework as well as our direct partners, ensuring that our solutions were feasible across all points of the spectrum of users.
One specific partner request was the need for an eligibility module within their application build. Credit Unions utilize these questions to verify their applicants during the decisioning process, with the approach varying across different unions.
Primary CTA is given to adding a conditional, but what about partners that only need individual questions?
Tethering conditional question creation so tightly to a dropdown option feels too "happy-path". How can we adjust entry points to accomodate all types of workflows?
Instead of dropping users into a module with a multitude of complex interactions, the revised design has them choose at the start how they would like to add eligibility questions, with visual cues indicating how the chosen layout would function. At any time, users can change their layout by restarting the module.
Though the interaction design of the conditional layout was largely kept the same, I added increased affordances to allow users to add, edit, and link their questions in the manner that they best understand. Users can freely add conditionals to the module and link them after building up their questions or can take the original approach of tethering a conditional directly to an answer to start.
Building on previously established patterns from modules like Custom Questions, the individual question layout allows users the option of up to two separate questions, with the user providing the content of those questions.
As part of their application build, users will be responsible for inputting the tiers of approval prices that applicants could be placed into.
Linear Stacking
Expand / Collapse Sections
Segmented Workflow
Product expectations were initially set on a maximum amount of pricing groups, that being ten, with the initial designs reflected this by loading in empty-state columns automatically. Though this accomplished the goal, it made the first section of the module feel extremely dense, especially for partners who may not have ten different groupings of prices.
I tightened the interaction design of the pricing columns, allowing users to add and remove columns as needed, allowing the design room to breathe instead of being packed with input fields.
I redesigned the original credit lines creation component to increase usability and overall understanding. Instead of relying on color states, clear affordances are presented, as well as the opportunity for users to automatically generate their lines, or to manually create them on their own.
This approach also matched other areas of our self-service product, that being the Pricing approach for our 2024 initiative.
Product Bookend
Though the bulk of the configuration pages were created, it was crucial to the UX of the product to create effective beginnings and ends to the products themselves. We created a simplified product creation and dashboard experience, designing an easily scalable "hub" for product information.
Users needed to be able to view their products at-a-glance in a simplified dashboard environment. To bolster the experience of the product and increase partner autonomy, we added in the functionality to freely create, duplicate, and delete products- removing the need for partners to utilize Amount as a confirmation funnel and imbuing trust into the process.
As a peripheral functionality, the dashboard also functions as an at-a-glance look at the state of their current product configurations, allowing partners to gain a valuable perspective on the implementation of their products.
Generating a new product is intentionally simple, and users are prompted to do as such from their dashboard. Once created, the product is added to their dashboard and is available for editing.
To bridge the gap between initial creation and the entirety of the portal, the product name and logo are automatically loaded into the "Branding" section of the product.
A common pattern noticed from client discussions is that, for the majority of partners, many of their configuration settings were shared across products.
To allow for a more seamless experience, users can easily "duplicate" one product, with its configurations being shared.
To avoid a dragging onboarding experience, the product loads "bite-sized" tips and directions about the admin portal, allowing users to freely jump in, explore, and set to building their product without being impeded.
Product Bookend
A crucial part of the Admin Portal was the review process prior to submitting configurations that had been entered. For the MVP, the submission process was treated as a final act, with no post-submission editing offered. If edits were needed, partners would need to contact Amount directly.
Because of this initial limitation, it became apparent that the submission process would need to consider two equally important use-cases:
With varying levels of complexity to the modules throughout the product, I needed to create a focused configuration summary process that allowed users to at-a-glance view their configurations, scaling from a granular to a broadened level.
Because the submission process was a particularly concrete action for partners, our post-submission product handling approach had to be accommodating and useful, with the option to contact Amount for edit requests directly baked in.
Module configurations are condensed into collapsed sections correlating to the navigation on the left. At a broad level, users can utilize the review page to get a glance of what modules are complete or incomplete prior to submitting.
Alternatively, users can expand individual sections to view the configurations that have been entered into that specific module. The information within these sections is intentionally pared down visually, relying on simplistic tables with minimal styling to allow for increased scannability.
After submitting configurations, partners can view the status of their submitted product through the main dashboard and can access the previous summary of their configurations for review.
If needed, partners can contact Amount directly through the "Request Edits" functionality, facilitating discussion and allowing their product configurations to be unlocked for further editing if required.
My final design consideration focused on evolving the experience of the portal to make the process of filling out modules easier to process at at-a-glance and, from a larger level, filling the potential gaps between partner understanding and module expectations through easy-to-access design resources.
Though partners could view module completion states at any time through the Review & Submit page, there was a noticeable lack of progress indication whilst working on modules, requiring users to repeatedly jump into and out of their review page in order to access and process information.
To bridge this UX gap, I redesigned the product navigation to add a secondary purpose in showing the status of the module itself.
Though the majority of partners had experience in the configurations of their product, we wanted to implement a way for users to easily access documentation and resources that Amount had previously set up for partners to utilize.
Because the Admin Portal was set to be launched to a relatively niche set of users (at first), the metrics set to track were meant to reflect Amount's first step towards a full-scale SaaS product.
Monitoring system downtime as well as overall system performance was crucial, as the Admin Portal was built on codebases that were slightly dated.
Through direct communication with partners, monitor overall satisfaction with the product and pinpoint areas of improvement from both a UX and technology perspective.
As the MVP continued to be used, monitor errors within modules of the portal itself in order to measure success and performance of the product.
Specific focus was meant to be put on tracking the overall time between configuration submission and implementation, and targeting future areas of improvement as Amount moved towards a full-scale SaaS product.
Though not directly design-related, the push towards a SaaS business model naturally intertwined itself with financial monitoring and the overall impact the Admin Portal could have on Amount's revenue streams.
The MVP of the Admin Portal launched shortly before the end of Q4 2023 for use internally. Prior to the launch, I worked closely with developers in daily QA sessions to walk through existing module builds, discuss any limitations that may have arisen, and discuss any compromises that may have been needed between the envisioned design and MVP possibilities.
I also set up plans for moderated testing with our external partners prior to the Version 1.0 of the Admin Portal. Though I departed Amount before this could be accomplished, I am confident that the existing MVP of the portal will continue to be expanded and improved upon with the design artifacts and guidelines I was fortunate enough to have set up.
LIAM MADIGAN
CHICAGO, IL
© 2024
LMADIGANDESIGN@GMAIL.COM