Introduction
It was a project partnering with IBM UX Team to evaluate the current product “IBM Developer” website and to propose solutions to improve the user experience.
User Researcher
Figma, Illustrator
Role
IBM Developer Website Evaluation
UX Research 2019
Tool
Overview
IBM Developer is a free web-based professional platform providing resources for software developers, IT professionals, and students worldwide to learn skills, communicate technologies and collaborate with each other. The IBM UX team has already completed an investigation into website performance and they found that the Net Promoter Score(NPS) showed a low satisfaction in the Community section. Our primary work was to assess the website’s current performance and understand users’ behavior, with a focus on community, from an external perspective.
Research Process
Interaction Map
The purpose to create this “ interaction map” is to better understand the information architecture, to learn about possible user flow, and to identify some structure problems of IBM Developer website.
Three problems regarding the current website’s structure:
-
No searching or sorting feature in the blog feature.
-
“All industries” option does not lead to more industries.
-
“Game Developer Meetup” does not show up when searching for game events.
Interview
Conducted four interviews with developers in order to learn about the typical needs and habits of developers, and in particular about how they interact with developer sites (not limited to IBM) and the broader developer community.
After the interviews, our group conducted interpretation sessions on interview notes, categorizing the findings, summarizing three key insights, and creating personas of three main types of users accordingly.
Insight of developers’ behavior:
Developers’ attitudes on the community:
-
Most users start with general web searches even if they are looking for answers on a particular site such as Stack Overflow, and they care more about the content of websites than the interfaces.
-
Users typically search for pre-existing answers to specific questions and rarely actively participate by posting their own questions or answers.
-
Developers generally only join communities or attend events directly related to their jobs, and rarely participate in online groups.
Comparative Analysis
Survey
Difficulty finding certain content
54
Low attendance rate for real-world events
Participants
Heuristic Evaluation
Following Jakob Nielsen’s 10 usability design heuristics, our group found that there were problems with consistency between different sections of the site and that certain key features had poor visibility on the main page and navigation menus, as well as that there were relatively few options for customer support.
Poor Visibility
Preference Test
While we only had 10 participants take the survey, our initial results were promising, with a majority of users responding that they felt that our redesigns would make the Developer site easier to navigate.
10
Criteria
Documentation and Help
10
Redesign site mockups and got 10 test results.
Usability Test
5
User Test
Find Forums
Find Event Info
Find a Tutorial
Locate Customer Support
Possible Solutions
Findings
1. Users are confused about the naming and placement of major features.
2. Users liked the content but found the overall structure of the site difficult to navigate.
3. Navigation (going back) of the website is poor.
4. There is no obvious way to contact customer support for help with using the site.
Recommendations
1. Retitle the external "Answers" link and add it to the main navigation bar.
2. Improve the discoverability of content and the overall structure of the website.
Site Search
Visual Consistency
3. Make sure every sub-page has an obvious link back to the Developer homepage.
4. Rename the "Feedback" page and add more customer support options.
Live Chat
FAQ & Contact Us
Form
Shortages of Research
-
The sample size for all of the steps was fairly small.
-
Aside from the user surveys, none of the participants we tested or interviewed had experience in using IBM Developer before.
-
​The participants of usability tests also skewed quite young, with most of them being in their early-to-mid 20s.
-
The tasks in usability tests were not particularly exhaustive of the site, and there were many features that we left untested as we were unsure how to incorporate them best.
Video
