With Love, Tife

With love, Tife

Email me

resume

SPIN OFF @ ALEX OYEBADE

Previous project

Next project

Takeaway

This project highlights the power of combining behavioral insights with thoughtful UX layering. By designing for actual user behaviour, we increased feature usefulness without compromising usability.

Conclusion

Tagmarks transforms bookmarks from a passive archive into a functional, user-curated knowledge base. The results validated our hypothesis that even lightweight organisational features can significantly enhance user productivity and retention.

Not included in this case study, but part of the final product and v2;

Tag management tools (rename, re-tag, delete).
Auto-tag suggestions based on tweet content or metadata (v2).
Personalized tag recommendations driven by Grok (v2).

Feedback and Expected Impact

We re-tested the updated flow with 20 participants, and recorded a 100% success rate in retrieving posts. The average retrieval time was 10s, and users preferred the toast interaction to the long-press.

After we launch the feature, we expect:

+60% reduction in retrieval time.
+25% increase in search-driven bookmark sessions.
+18 NPS lift among participants who used tags

Design – Iteration II (Relevant Screens)

The Bookmarks screen now reflects the tags, and tags can be selected via the dropdown.

Design – Iteration II

In the second iteration, we moved away from the long-press gesture and instead adopted a toast-based prompt.

What changed

A toast message inviting the user to add a tag.
A dropdown menu to let users select/search an available tag or create a new tag.

Why this worked

The single-tap bookmark behavior is preserved.
Tagging becomes progressive.
Improved discoverability and flexibility for all user types.

Design – Iteration I

We introduced a long-press gesture on the bookmark icon (2-second hold) to trigger a tag input dialog.

What worked

Clear modal interface for entering tags.
Tag filtering improved post retrieval significantly (see iteration II).

What didn’t

Users had to learn a new interaction (long-press) to add tags.
No ability to save posts to existing tags directly.

The Results

We found out that 80% of users found it difficult to find bookmarks older than 1 week, with majority dropping off within 15 seconds of trying to find a post. 32% resorted to Google search as a workaround.

We also found out that 100% of participants found tagging helpful in a prototype, and wanted bookmarks/tags to be searchable and filterable.

Research Study

To validate the problem and assess potential solutions, we conducted a mixed-method research study involving 100 high-frequency bookmark users in 10 countries (US, Japan, India, Indonesia, UK, Brazil, Turkey, Mexico, France, and Saudi Arabia). We gained qualitative insights on retrieval attempts via task-based usability testing and open-ended interviews to explore behaviors and expectations.

Our aim was simple. Find out:

Where and why do users drop off when retrieving bookmarks?
How do they currently work around limitations?
What mental model do users expect for organizing saved content?
Would tagging increase retrieval success?

PREVIOUS Bookmark User Flow

Limitations: No labelling, tagging, or filtering available.

Retrieval: Users must scroll or use keyword search, which is often ineffective due to vague memory or post content.

THE CHALLENGE

Design a seamless and efficient workflow for users to saved, organise and retrieve bookmarks.

BACKGROUND

Every minute, users on X (formerly Twitter) bookmark over 500,000 posts — from threads to memes and educational content. Despite this, the feature remains under-utilised in practice with 80% of users abandon their search when trying to retrieve an older bookmark.

The reason? Bookmarks exist as a flat, reverse-chronological list, offering no way to categorise or filter content at scale. In many cases, users Google keywords instead of using X’s built-in multi-level search.

MY PRETTY LITTLE DELUSION

As a Senior Designer at X, I led the end-to-end design of Tagmarks, a feature that lets users organise saved posts with custom tags for faster, smarter retrieval.

Through generative research and iterative prototyping, we launched a solution that reduces search time by 60% and increased daily active bookmark sessions by 25%.