AdSuite is a Saas project of a Bulgarian/American startup that converges multiple end-to-end business processes of the programmatic advertising business which includes automation and billing and financial services, campaign tracking and trafficking, and other apps like CRM, monitoring & performance analysis, reporting tools, dashboards, etc.
I joined the company in 2022 and my job was to identify problem spaces with the current state of the web apps. It was a pivotal moment as the company was about to unleash its first MVP to the client. At that point, the company was cruising on code but lacked a solid UX design game plan, with no established design role in their processes. During that time I was involved in multiple projects and activities, that enriched my experience in research, visual and experience design, agile processes, and cross-functional team collaboration as well.
Error Handling of Recipient Errors
Designing a PDF Generator Feature and Document delivery services revealed critical recipient errors in the automated operations. The system exhibited issues such as contact data lacking critical information for the business flows such as valid emails or physical addresses, posing a challenge for automatic document delivery service.
My process included participation in a discovery workshop with the technical team to uncover the edge cases on various sources, diagramming and problem statements, solution architecture, user flows, and design of effective error prevention in the UI on multiple levels.
I have conducted a thorough deep exploration with the technical teams about the data we receive from integrations and I traced the journey of the contacts through manual and automated operations at all system levels, so I can track the errors and find the cases where we can handle it.
This risk especially concerning given the high volume of operations planned for execution in our Billing Engine and Document delivery service, handling nearly 100,000 monthly invoices per client, with approximately 5% affected by recipient errors, which is equal to approx. 5000 invoices per month with recipient errors and that system condition risked expensive errors, including delivering to incorrect addresses or using unvalidated USPS physical addresses for paper deliveries.
Identifying problem areas & UX Prioritization
During the MVP delivery, I conducted a series of usability tests and client interviews, supplementing these with SUS testing to gather authentic user data and unlock data-driven design decisions for the next phases.
Methods
1. Usability Tests and Client Interviews
Executing a total of 22 usability tests uncovered the product usability score, inconsistencies, functionality issues, and business gaps between expectations and results, and also new opportunities for business growth of the Saas product. Engaged in client interviews to align design goals with business expectations. This instrument helped me to demonstrate how the design can solve the functional problems of collecting authentic data and provide opportunity areas for growth and development for my team.
2. Supplementing with SUS Testing
Implemented System Usability Scale (SUS) testing to quantitatively measure the product's usability score and a key metric for iterative usability score testing. The SUS scores provided a baseline for evaluating improvements over subsequent iterations.
Having everything I learned from qualitative research and user interviews I worked on extracting, analyzing, and documenting the obtained results in reports, creating a Research repository, and summarizing and presenting them to stakeholders. My research and insights contributed to the product roadmap, prioritization, and planning as we worked together with the Product Owners to include them in the product backlog.
I collected sufficient data to create the ideal user persona method which helped our team to understand better why we build what we are developing and who is going to use it afterward in general this knowledge enhanced the process of writing better user stories and Jira tickets execution.
Improvement of validation errors in modal forms and input fields
The existing system faced challenges related to validation errors in forms and input fields within 28 modal forms, particularly with higher severity scores, inadequate input validation, and poor field labeling. The system lacked robust error prevention mechanisms, leading to a poor user experience. The task at hand was to research the current Front-end (FE) validations we can technically maintain and study the collected data from the usability tests, identify shortcomings, and collaborate with the FE team to enhance the design of forms and field validations. So the outcomes of my work were:
1. Improved Input Validation: Enhanced the input validation mechanisms to detect and prevent errors effectively. Implemented real-time validation, offering instant feedback to users.
2. Clear Field Labeling: Revamped field labeling to be concise, descriptive, and consistently positioned. Utilized clear and accessible labels to guide users through the form fields, reducing ambiguity and the likelihood of validation errors.
3. Improved Placeholder text in the fields to navigate the user intuitively to the task.
Before
After
Reducing Friction on Critical user actions
Some of the double-confirmation dialogues were part of business processes, other was the idea of a messy development process, which resulted in inconsistent and problematic double-confirmation dialogues. To address this challenge, I analyzed user feedback and engaged in collaborative discussions with stakeholders to comprehensively understand the pain points users experienced and understand better their processes, use cases, and scenarios within the platform. I crafted a solution that involved a thoughtful redesign of confirmation dialogs, following Interaction design principles, prioritizing clarity, simplicity, and consistency, using a systematic approach, and implementing color coding and icons, action triggers, and button CTAs.
Systematic research of components
New confirmation mockup
PDF Generator UX/UI Optimization
The PDF Generator is a service that automates creating PDF documents on billed orders and it was implemented with no users in mind. Unfortunately, many usability issues hindered user performance, leading to dissatisfaction and clients needed better solutions to it. Users faced difficulties in understanding the current state of the system during PDF generation and they found confusing the feature. My role as a designer was to improve the usability score of this feature and implement interaction design principles like the visibility of system status and error prevention and make the feature easier to use.
My process included a deep exploration of the business processes' level and technical functionality. I designed a user flow with different states of the service - entry state, progress state, error, and success state. Created visual design improvements, proposals, and development hand-offs.
Search Component - UX/UI Improvement
Another usability-reported issue was the Global search functionality with several errors in consistency, quality of information, and visual patterns. I researched the specific data display opportunities with both of the customers and validated the possibilities of implementing them with technical teams in my design I focused on designing clear component states and visible and memorable interfaces, inspired by the Atlassian design system. I exchanged the control selectors that are least talkable for our users with tabs, so they can navigate through different search databases. I aligned the visual design on all of the screens and filled them with object name data.
Before
After
Improving User-Centricity in USPS API Integration
The initial implementation of the USPS API integration suffered from a lack of user-centric focus, resulting in various problems. These issues ranged from user confusion to inefficient processes, ultimately impacting the effectiveness of the integrated feature. The USPS API responses were thoroughly examined, and a deep dive into the intended functionality at the business level was conducted.
