Revisiting Tweets Made Easy: Twitter Revamp

Product Design
Since launch, Twitter has had billions of Tweets posted (with millions more being posted every day), providing a constant flood of new posts and content. With this constant stream of new Tweets, sometimes re-finding a Tweet you want to share with a friend is difficult to locate even minutes after discovering it.

Objective
Giving users more ways to filter through the countless Tweets on a feed, effectively shortening the time a user would need to spend looking for a Tweet they wanted to share or rediscover.
‍
Long-term goal
Making Twitter an overall better and more accessible platform. This led to me work on making the search process for Tweets to be easier and smoother.
‍
Role
I led the design of the search and filter revamps on different Twitter pages.
Overview
Scope      Interaction and Design
Platform iOS
I conducted interviews with frequent users of Twitter to get their insight on how they look for old Tweets they want to find again.
Research
The key takeaway I got from these interviews is that the process of rediscovering a Tweet is a major pain point, which takes users anywhere from 5 minutes to 1 hour to find, if they don't give up before then.
Twitter's advanced search function is not used often, due to being hidden away in the search results window and its unappealing appearance.

From here, I decided to focus on redesigning Twitter's search function in some capacity. These conversations led to me to the following opportunities:
How might we create an easier way to revisit tweets?
How might we make the search process smoother and faster?
How might we create a more visually forward Twitter experience?
Map
Below is an overview of the experience as it stands today, identifying the step of using the search function as the focus for this project because I found that changing this feature would lead into different improvements that could be made across Twitter as a whole.
I focused on how we could make a smoother and faster search process because as stated above the map,  I found that changing this feature would lead into different improvements that could be made across Twitter as a whole. This solution led me to sketching out how searching could be improved on different kinds of screens.
8 low-fidelity sketch ideas. I considered different screens for search options.
The idea I decided to go with in the end. I sketched out ideas for searching profiles and bookmarks along with the normal search option.
The runner-up solution resembled Facebook's and Instagram's activity logs.
Exploring Solutions
After choosing a direction, I set out to more explicitly define design goals and success criteria:
By adding more search functionality on Twitter, users will have an easier time keeping track of Tweets they want to refer back to.
The search changes on different Twitter pages will fit in with Twitter’s current design and will not feel disruptive or out of place.
Users spend less time looking for Tweets they want to find with the additional search functionality.
Storyboard
My user’s journey began with a user looking for a Tweet they wanted to show their friends and ended with them finding the content they want to share with ease. Other important moments included:
‍
‍1. the search function featuring new buttons that can toggle different criteria (by post or people)
2. interactive buttons that can include and exclude certain posts
Highlight 1: 
Search buttons that can help filter posts more easily than Advanced Search
Highlight 2:
Search filters include and exclude different types of posts to show
Prototyping a
Successful Solution
Prototyping a
Successful Solution
Prototype
I generated a realistic prototype to bring my solution to life and get high-quality feedback from users with Figma. The prototype features three different areas I felt like searching on Twitter could be improved: the search section, the profile, and the Bookmarks section.
Try out my prototype here!
This YouTube video presents a click-through of my Webtoon Figma Prototype, displaying all the screens created to bring this idea to life.
What changed: The search filters have a more interactive flair, allowing users to decide what they want to see better; posts can be toggled by top or recent.
What changed: by adding a search button to profiles, users can search directly on any profile using keywords and filtering with the tabs Twitter profiles already contain (Tweets, Tweets & replies, etc).
What changed: Bookmarks' functionality improved greatly, enabling search similar to profiles and the search tab. Users can search by keyword, and filter by post type or people.
I identified target users through my personal contacts and assembled a research script to ensure I got high-quality & unbiased feedback. Interviews were about 45-60 minutes long via video call.

Each user was given the task of going through the motions of searching for a Tweet they wanted to share with a friend.

While I had to guide my users through the process of discovering the profile and Bookmarks screens for new features, all of the users I interviewed were impressed with the changes made to the search tab. One user stated, "This layout seems easier to use than using Twitter's advanced search for sure."
User Testing & Evaluating

Key Takeaways

After completing interviews, I revisited my notes to find large patterns and inform next steps.
Some users remarked that the search filters did not work the way they initially thought.
For my next iteration, I will make the "checkmark" filter only show the indicated field (i.e. if "Photo" is checkmarked, only Photos will show up in search results).
The most valued search experience change was to the Bookmarks tab.
Many of my users were shocked to discover that, currently on Twitter, the only feature you can do in Bookmarks is to erase them. They were thrilled to see my change, which utilizes the list better.
Working with Twitter's existing features made the redesigns appear familiar and trustworthy.
By replicating Twitter's own search bar and using Twitter-esque elements across the three screens I changed, users were quick to understand what each function did and how to use it.