I have conducted several client interviews as part of my primary qualitative research. These interviews were valuable in identifying the key pain points and challenges experienced by the users.

• Research subjects: 9 loan structuring analysts
• Research Methods: Interviews, product prototype testing

Target User Analysis

Loan structuring analysts often need to compare loan term sheet files and loan agreement files to make sure that all terms in the term sheet are included in the agreement. This can be a time-consuming and labor-intensive process, as it requires manual search and comparison.

According to interviews, users would like a system that can automatically extract terms from the term sheet and locate corresponding terms in the agreement. This would streamline the workflow and reduce error rates.

Current Loan Document Review Process: A Tedious and Error-Prone Workflow

The steps users typically take to compare loan documents:

1. Collect digital versions of the term sheet and agreement

2. Find all involved terms in the term sheet and read them one by one

3. Manually search the agreement for each term and find the corresponding term

4. Carefully compare the details of each term to confirm consistency

5. Record the final review results and opinions

1. Automatically extract all involved terms from the term sheet
2. Precisely locate each corresponding term in the loan agreement
3. Highlight the differences and inconsistencies between terms
4. Have text and visual output for easy user review
5. Simple and easy to learn
6. Support common file formats such as PDF, Word

After the user research, we designed this streamlined flow to address the specific pain points faced by loan analysts. Each step targets the frustrations they expressed, aiming to make their work more efficient, and accurate.

After understanding the product goals and user needs, I first quickly sketched out possible layouts and content blocks for key pages. This allowed me to freely think through different structural combinations and arrive at a reasonable initial concept.

By rapidly sketching lo-fi wireframes, I was able to thoroughly explore different page relationships and layout options, and quickly make changes and test them out. This ultimately yielded a layout approach that the team collectively found reasonable, laying the foundation for high-fidelity prototyping down the road.

Designing proper information architecture is crucial for a product’s user experience. A logical information structure can make complex content more organized so users can quickly find needed information without getting lost in trivial details.

For this project, I first needed to understand the logical relationships between key information pieces – terms are extracted from uploaded files, and comparison results are generated by matching terms with the loan agreement file. There are clear hierarchical and sequential connections between the information.

Using the completed set of digital wireframes, I created a clickable prototype. The primary user flow I created was viewing the results of terms, so the prototype could be used in a usability study.

Research goals

1. Evaluate the usability and effectiveness of the UX flow
2. Evaluate the understandability of workflows like file upload and term extraction
3. Test usability of term comparison and result presentation functions
4. Discover issues and improvement opportunities in interface elements
5. Determine user satisfaction and gather feedback for improvements.

Participants

1. Loan Structure Analysts: 3 participants
2. Loan Approvals Officers: 2 participants

Testing tasks

1. Create a review task
2. Compare terms against the agreement file
3. Verify the comparison results and highlight the correct clauses in the loan agreement file
4. Annotate term sheets and loan agreement files manually

Data Collection

1. Record task time and steps taken
2. Note down problems encountered in flows
3. Gather subjective feedback on interfaces and interactions

Results Analysis

1. Analyze task success rates and workflow usability
2. Identify problematic interface elements and solutions
3. Compile feedback to inform optimization

Testing KPIs

1. Task success rate
2. Task time
3. User Satisfaction

Results

• 4 users encountered difficulties during the term comparison task, with only a 50% complete success rate.
• 4 users believe that a save feature is needed so that they can review the work in parts instead of all at once.
• 4 users have expressed that the design of the comparison document is not intuitive.