Improving Experience for
PyKX Documentation

Improving Experience for PyKX Documentation

Improving Experience
PyKX Documentation

Simplified technical documentation navigation so developers can find information faster, get started with confidence, and move through content more easily.

Software & Tech

Team Project

2024

Software & Tech

Team Project

2024

Challenge

The PyKX documentation grew over time and became hard to navigate. Developers often struggled to find what they needed, know where to start, or move between high-level ideas and real examples.

Design Goal

Restructure the information architecture. My focus was on creating a logical flow that supports both high-level conceptual learning and rapid, hands-on implementation.

Outcome

The redesign significantly improved the experience, achieving a 70% increase in task success, 8x faster completion, and a 40% reduction in dead clicks.

My Role

UX Designer & Researcher (Collaborated with Senior Product Designer)

User Research

Prototyping

Wireframing

Card Sorting

A/B Testing

Data Synthesis

The Challenge

The PyKX documentation grew over time and became hard to navigate. Developers often struggled to find what they needed, know where to start, or move between high-level ideas and real examples.

Design Goal

Restructure the information architecture. My focus was on creating a logical flow that supports both high-level conceptual learning and rapid, hands-on implementation.

Outcome

The redesign significantly improved the experience, achieving a 70% increase in task success, 8x faster completion, and a 40% reduction in dead clicks.

My Role

UX Designer & Researcher (Collaborated with Senior Product Designer)

Card sorting

Navigation Testing

Prototyping

Wireframing

Usability Research

A/B Testing

Data Synthesis

Results

Impact and Outcomes

Impact and Outcomes

70%

Significantly Improved

Task Success Increase

Users are completing core tasks with significantly fewer errors.

3-8x

From 1min to 7sec

Faster Completion

Streamlined workflows have dramatically reduced the time needed to finish complex actions.

40%

Reduced Friction

Reduced Dead Clicks

Enhanced clarity and feedback have eliminated friction points and improved user confidence.

70%

Significantly Improved

Task Success Increase

Users are completing core tasks with significantly fewer errors.

8x

From 1min to 7sec

Faster Completion

Streamlined workflows have dramatically reduced the time needed to finish complex actions.

40%

Reduced Friction

Reduced Dead Clicks

Enhanced clarity and feedback have eliminated friction points and improved user confidence.

Research

Understanding the Users

Understanding the Users

A deep dive into user behavior and pain points to inform the new architecture through quantitative and qualitative analysis.

Three Core Research Pillars

User Interviews

Reviewed feedback and grouped common pain points like onboarding hurdles, unclear terminology and structure.

Card Sorting

Conducted open and closed sorting sessions to understand mental models, labels, and how to group topics effectively.

Navigation Testing

Tested existing structures with real-world tasks to identify specific friction points and exactly where users got lost.

Three Core Research Pillars

User Interviews

User Interviews

Reviewed 12 user interviews to uncover friction points in the onboarding journey and clarify confusing terminology.

Card Sorting

Card Sorting

Synthesized 150+ sticky notes from open and closed sorting sessions to align the site architecture with user mental models.

Navigation Testing

Navigation Testing

Validated the hierarchy using Maze, identifying exactly where users experienced friction during real-world tasks.

From Raw Data to Information Hierarchy

From Raw Data to Information Hierarchy

By analyzing user-suggested categories, I restructured the documentation into three high-priority themes: setup, reference, and troubleshooting.

Installation

28 Initial Cards

"I expected a Quick Start guide right here under installation."

User #04 โ€ข Senior Dev

"Took 30 mins to install, I don't have time for this"

User #12 โ€ข DevOps Eng

Documentation

64 Initial Cards

"I wasn't sure if the API reference belonged under Documentation or a separate Reference tab."

User #09 โ€ข Frontend Architect

"Dummy tables could be useful to very quickly give 1000 foot view to apply a certain functionality."

User #15 โ€ข UI Developer

Errors

58 Initial Cards

"Error codes are useless if I can't find a direct mapping to the fix."

User #02 โ€ข Fullstack Dev

"Common issues should be grouped by the stage of development."

User #21 โ€ข QA Lead

Installation

28 Initial Cards

"I expected a Quick Start guide right here under installation."

User #04 โ€ข Senior Dev

"Took 30 mins to install, I don't have time for this"

User #12 โ€ข DevOps Eng

Documentation

64 Initial Cards

"I wasn't sure if the API reference belonged under Documentation or a separate Reference tab."

User #09 โ€ข Frontend Architect

"Dummy tables could be useful to very quickly give 1000 foot view to apply a certain functionality."

User #15 โ€ข UI Developer

Errors

58 Initial Cards

"Error codes are useless if I can't find a direct mapping to the fix."

User #02 โ€ข Fullstack Dev

"Common issues should be grouped by the stage of development."

User #21 โ€ข QA Lead

Installation

28 Initial Cards

"I expected a Quick Start guide right here under installation."

User #04 โ€ข Senior Dev

"Took 30 mins to install, I don't have time for this"

User #12 โ€ข DevOps Eng

Documentation

64 Initial Cards

"I wasn't sure if the API reference belonged under Documentation or a separate Reference tab."

User #09 โ€ข Frontend Architect

"Dummy tables could be useful to very quickly give 1000 foot view to apply a certain functionality."

User #15 โ€ข UI Developer

Errors

58 Initial Cards

"Error codes are useless if I can't find a direct mapping to the fix."

User #02 โ€ข Fullstack Dev

"Common issues should be grouped by the stage of development."

User #21 โ€ข QA Lead

What We Learned

What We Learned

From research and mapping, we surfaced three core usability issues:

Navigation Lacked Structure

Users consistently struggled to find specific documentation topics. Existing labels were too broad or technically jargon-heavy.

Confusing Terminology

Significant overlap existed between "Reference" and "How-To" sections, causing cognitive load and decision paralysis.

Poor Content Discovery

There was no clear linear journey for new users, leaving them stuck between initial setup and advanced learning paths.

Navigation Lacked Structure

Navigation Lacked Structure

Users consistently struggled to find specific documentation topics. Existing labels were too broad or technically jargon-heavy.

Confusing Terminology

Confusing Terminology

Significant overlap existed between "Reference" and "How-To" sections, causing cognitive load and decision paralysis.

Poor Content Discovery

Poor Content Discovery

There was no clear linear journey for new users, leaving them stuck between initial setup and advanced learning paths.

These insights directly shaped the new site structure and menu design.

Design & Test

Prototypes & Validation

Prototypes & Validation

Propotypes

Propotypes

Working closely with the Senior Designer, I contributed to creating testing wireframes and prototypes for the new documentation layout. My focus areas included:

  • Prototyping alternative navigation models

  • Documenting page relationships and navigation flows

  • Testing and iterating on sidebar layouts and in-page links

Before

An overcrowded left navigation forced users to guess where information lived.

An overcrowded left navigation forced users to guess where information lived.


After

Grouped and renamed sections improved clarity, with content nested under more intuitive categories

A/B Testing

To measure the impact of the new navigation, we ran an A/B test in Maze with 60 participants, comparing the redesigned layout with the original navigation. Participants were randomly assigned to reduce bias and completed 10 tasks across both versions.

๐Ÿ“Š Metrics Measured:

  • Success rate

  • Time to completion

  • Misclicks

  • User confidence and feedback

๐Ÿ”ฅ Heatmap Insights

To better understand user behavior, we reviewed click heatmaps from Maze for both the original and redesigned navigation.

Before

Clicks were scattered, showing confusion about where to go.

After

Clicks followed a clear path, showing more confident navigation.

Reflection

What I Learned

What I Learned

This project was a significant milestone where I am proud to have successfully translated 150+ data points into a high-performing navigation system. Seeing my research directly result in measurable improvements to task success and time-on-task solidified my confidence in creating impactful user experiences.

#1

Validating navigation updates through card sorting and A/B experiments

#2

Linking qualitative findings with numerical data

#3

Evaluating design effectiveness via task success, completion time, and user feedback

#1

Validating navigation updates through card sorting and A/B experiments

#2

Linking qualitative findings with numerical data

#3

Evaluating design effectiveness via task success, completion time, and user feedback

Thank you for visiting! ๐Ÿ‘‹๐Ÿผ

Thank you for visiting! ๐Ÿ‘‹๐Ÿผ

Thank you for visiting! ๐Ÿ‘‹๐Ÿผ

Thank you for visiting! ๐Ÿ‘‹๐Ÿผ

Thank you for visiting! ๐Ÿ‘‹๐Ÿผ