Some customers were feeling ignored or dismissed when a bug fix that was launched did not solve their problem.
Better communication with customers
Ability to track number of customers still blocked
Improve typography/visual design of comment threading
Improve NSAT score by 10 points
Lead design in collaboration with PM/developer
This project was part of a 3 part plan to improve net satisfaction for customers when bugs are resolved. Areas that needed the most improvement were:
Clarity of prioritization of problems (bugs)
Timely resolution of problems
Searchability of solutions
This project focused on clarifying prioritization of problems, and improving communication as it pertains to closed/fixed states.
Before designing the resolution reaction component, I suggested improvements to the comment threading.
Improve typography, flatten comment threads
Make better distinction between solutions and comments, better container for "best solution" or fixed state.
Customer can hide comments to go straight to solution section below
Each solution and its comment threads separated. More accessible visual treatment.
Before going to Figma I find it helps to sketch out what the stakeholders are asking for (and my ideas for that) to see if it raises more questions.
Slide 1: There wasn't a clear reason to offer the customer so many choices to react to the question. I also wanted information on whether/how customers use the share buttons in the forum platform they were using.
Slide 2: This sketch was to drive the discussion about how the "reaction buttons" would work with comments. There was debate over whether to require comments for "No" answers to get details on why/how the fix didn't work.
Slide 3: Considerations for requiring comments with this many options
Slide 1: With 4 choices - this version was wireframed to show that scenario
Slide 2: With 3 choices - decided to keep alternate solutions in solution section of page
Slide 3: With 2 choices + "respond all" component I designed in the backend UI. Developers needed a way to comment to the entire group at once.
Slide 4: With 2 choices - After live testing with 3 choices, stakeholders decided they preferred basic numbers on whether it worked or not.
The NSAT score did improve by 12 points. Microsoft employees spend about 30% less time responding to customer complaints, now that they don't have to field them one at a time.
The Resolution reaction component was so useful it was added to other closed states, such as not a bug, need more info, duplicate, and low priority.
Clicking "Yes" auto generates a yes comment on the thread.
Clicking "Yes" auto generates a yes comment on the thread.
Clicking "No" opens the comment box for the customer to give details. If there's no comment, the "No" vote is not counted.
See new Developer Community site with feedback tool still in use here
B Blakesley ©2024
