team
4 researchers
Role
UX Researcher
duration
3 Months
Tools & Methods
Atlas.ti, FigJam, Observational Study, User Interviews
Investigating digital clutter on social media
A discovery study on how people save and bookmark content on social media, what makes it pile up, and how to mitigate it.
overview
About the project
Over three months, my team of four ran an observational study and a round of user interviews to better understand users' experiences with saving, bookmarking, and sharing social media posts, and how to prevent clutter.
We each conducted two of the eight observations and two of the eight interviews, and collaborated on study design, coding, synthesis, and recommendations. This was a discovery study, so the deliverable is the research itself: a synthesized set of behavioral insights and a prioritized list of design implications.
Highlights from the study and process are detailed below, and the full report can be found here.
Research
Why digital clutter on social media is worth studying
02
existing research zooms out too far
Existing research treats clutter as one big bucket and rarely zooms into specific spaces like social media, where saving works very differently from email or files.
03
existing solutions are reactionary, not preventative
Products like Unroll.me clean clutter up after it piles up. Almost nothing looks at how and why it accumulates in the first place.
How can a technology-based solution help mitigate the accumulation of digital clutter on social media?
We coded transcripts in Atlas.ti, built affinity diagrams in FigJam, and translated the themes into persona spectrums, personas, journey maps, and a priority matrix.
My part: two observations, two interviews, and shared work on coding, synthesis, and recommendations.
Insights
The core patterns explaining user preferences and habits
We summarized our findings into four categories, subsequently supporting the following synthesis:
Insight 01
saving is social
Every participant saved with other people in mind — saving to share later, or sending a post to a friend as a way of saving it. So retrieval doesn't just live in folders; it lives in DMs too. Any solution (like search) has to reach saved messages, not just saved collections.
Insight 02
convenience is a priority, but varies by user type
Everyone optimized for convenience; they just did it differently. High-effort organizers built detailed folders for easy long-term retrieval, while low-effort savers dumped everything into a catch-all for instant convenience. That pointed to different supports — tagging for organizers, AI-suggested folders and date filters for catch-all savers — plus search and duplicate cleanup, which both groups wanted.
Insight 03
content type drives retrieval behavior
Users saved three kinds of content — promotional (products, events, restaurants), tutorial (recipes, DIYs), and inspirational (ideas, things they like) — and engaged with each differently. Active intent (promotional/tutorial) led to longer, action-oriented sessions; passive intent (inspirational) led to quick, nostalgic browsing. That suggested nudges and context indicators to help people act on saved items, and a highlight reel to resurface buried inspiration.
Design Implications
Where the research points the design
We mapped every idea against impact and feasibility in a priority matrix. The clearest wins:
Search: the single most-wanted feature; two participants named it before we even asked.
Date filters: All participants leaned on chronology to find old saves.
AI-folders & duplicate cleanup: low-effort organization for catch-all savers.
Tagging: structure for high-effort organizers.
Nudges: motivate users to actually try saved promotional and tutorial content.
Priority Matrix
Feature
Category
Priority
Impact
Feasibility
Search
Retrieval
Highest
High
High
Nudge to try saved content
Trying content
Medium
Medium
Medium-Low
Nudge to clean up content
Organization
Low
High
Medium
Highlight reel of old saved content
Retrieval
Low
Medium
High
Filter
Retrieval
High
High
High
AI Folders
Organization
Medium
High
Medium
Time Commitment Indicator for DIY
Trying content
High
Medium
Low
Effort Indicator for DIY
Trying content
High
Medium
Low
Reviews/Popularity rating for restaurants
Trying content
Medium
Medium
Low
Tags for saved content
Trying content
High
High
High
Deleting duplicates
Organization
Medium
Medium
High
Closing thoughts
What I learned
01
The value of personalization
"Convenience" was universal, but it pulled in opposite directions: power-organizers wanted more structure (tags), while catch-all savers wanted less effort (AI-suggested folders). This was a good reminder that you can't design to a stated value; you design to how different people actually enact it.
Additionally, half of participants shrugged off their untouched saves, noting that "it's just social media." That pushed back on our own premise and steered the recommendations toward optional, personalized support rather than a one-size corrective – closer to reducing friction than enforcing tidiness.
What's next?
01
Additional testing
Our sample skewed heavily toward Instagram and toward a narrow demographic (recruited through one university participant pool), so the findings lean exploratory rather than generalizable.
02
Prototyping
Turn the implications into a prototype: implement the top priority-matrix features to test whether they actually reduce saved-content clutter.
