### Background
The Explore feed is the first thing readers see when they open the Wikipedia app and for years, it hasn't been doing its job.
Engagement is low, the card layout feels static compared to modern feeds, and many readers don't realize it can be personalized. Meanwhile, Wikipedia's depth and breadth — its greatest strengths — can feel overwhelming without clearer structure and relevance signals.
We redesigned the Explore feed into a a two-tab experience called "Home", where 'Community' surfaces Wikipedia's best editorial content in a familiar, scannable format, and For You delivers a full-screen, immersive experience personalized to your interests, reading history, and location.
This is not a social feed and everything surfaces from Wikipedia's own articles and community.
The "Home" feed will soon be released on the Android app, and now we are beginning to scale it to the iOS app.
#### Strategic Alignment
This work directly supports the Consumer Strategy goal of reaching new audiences and addresses our biggest retention gap: the step from new reader to someone who comes back regularly. It maps to the annual plan KR to perform at least one experiment per platform showing a practically significant improvement in logged-out casual reader retention or an indicator metric.
###What we learned before building
Before designing, we ran two studies to understand what readers actually need from the feed.
**In-app survey · December 2025 · 1,184 respondents**
Current app users universally want more personalization and more interactive content — games, quizzes, and content tied to their interests and reading history. A significant share weren't aware the feed could be customized at all, despite it being the first tab they see on every app open. Games was the least-used feature in the feed and simultaneously one of the most requested improvements, signaling a major discoverability gap.
**Qualitative interviews · December 2025 · New and non-app users**
Participants brought strong expectations from other feed-based apps. When the Explore feed lacked familiar relevance cues — recency, popularity, personal connection — it felt hard to evaluate and easy to leave. Low-effort, remixed content performed best: features like On This Day resonated because they feel approachable and finite. Participants wanted personalization without algorithmic entrapment.
**Prototype testing · Early 2026 · 14 participants**
We tested two feed directions with new and casual readers: Focus (full-screen immersive) and Streams (scrollable card feed). The split was exactly 50/50 on preference — but for complementary reasons. Focus felt modern and conducive for daily use; Streams felt organized and credible. That finding directly shaped the final design: the Community tab uses the Streams approach, and the For You tab uses the Focus approach.
"It gives you that swiping gesture which is really addicting... this is a good way to do it because people are so used to this swiping gesture." — Participant, prototype study
"I would probably open it again just because I like to read something in the mornings before going into social media... just a little bit more substantial or meaningful." — Participant, prototype study
###The Experience
###Two Tabs, Two Experiences
The redesigned feed has two tabs accessible from the top of the Explore screen:
**Community**
A scrollable feed of content that exposes curation and information about the broader community and movement: In the News, Featured Article, On This Day, Top Read, Picture of the Day, and more. Familiar, structured, and updated daily.
**For You**
A full-screen, swipeable experience built around you. Each module is a vertical stop in your feed; within each module, you swipe horizontally through a stack of cards. Modules include:
- Article Recommendations based on your interests
- Because You Read — articles related to what you last read
- Continue Reading — your most recent article and your reading list
- Places of Interest — Wikipedia articles about places near you
- Discover Recommendations — the latest recommendations based on your reading list
- Random Article — one surprising article, shake to shuffle
- Games — a teaser question from today's game with a path to play more
At the end of the feed, users see a "That's it for today" card with the option to add interests, visit the Community tab, or loop back to the top. In the future this screen could serve as a screen of the daily digest or a quiz of what was in the feed that day.
####Onboarding
On first meaningful use, readers are invited to select topics they care about. The flow reuses elements of the Android reading list onboarding pattern — clear visual tiles, and search by topic or article. It's lightweight and skippable. Readers also choose their default landing tab (Community or For You), so they arrive at the content most relevant to them from the first session.
#### Personalization
Personalization in the For You feed is transparent and user-directed. Every card tells you why you're seeing it: "Because you follow Arts", "Based on your reading history", "Places of interest". Readers can dismiss content, save it, hide modules, and update their interests at any time from Settings → Explore Feed → What's driving your feed. No black-box profiling. No algorithmic entrapment.
###Design Principles
- Retention before depth — reasons to come back matter more than any single deep read
- Discovery through remixing — snippets and previews lead into articles, never replace them
- Transparent personalization — "Because you follow..." not a black box
- Calm, trustworthy browsing — productive and credible
###Core Requirements
#### The feed must
- Be immediately useful on app open
- Surface a mix of fresh content (what's new), familiar content (topics the reader cares about), and community content (Main Page)
- Clearly signal freshness — timestamps, "trending", "new this week" language where appropriate
- Allow readers to save, share, dismiss, and hide content — all actions reversible
- Support the full range of Wikipedia's languages as configured by the reader
#### The feed must not
- Summarize content using LLMs — all content comes from Wikipedia articles
- Function as a social feed with likes, comments, or follower counts
- Require account creation
- Feel unstructured or like a catch-all for unrelated feature
####Onboarding must
- Offer popular and trending topic suggestions
- Allow search by topic or seed article
- Be lightweight and skippable — no forced setup
- Let readers set their default landing tab
### How we'll know its working
####Key Result
DE3.1
By the end of Q2, achieve a 0.5pp increase in logged-out 21-day retention on mobile web and a 1pp increase in logged-out second-week return on apps, for readers presented with updated mobile experiences compared to readers in the control group
#### Target Reader
Someone who installs or opens the Wikipedia app with intent to read a specific article but has no established habit of returning — a new reader building their first relationship with Wikipedia.
### Release
We are aiming to release as an A/B test by August 15, 2026
- A= Explore, control
- B= Home, variant
#### Primary Hypothesis
If we scale the new Home Feed from Android to iOS we'll see a 8% increase logged-out second-week return for users in the variant group with access to Home compared to control group with access to Explore, as measured by an A/B test.
#### Metrics
####Key results
- 1.1 Key Metric: 8% increase in logged-out second-week return for users in the variant group compared to control group
#####Guardrails
- 2.1 No more than 3% of users turn off all modules and leave them off over course of experiment
- 2.2 No more than 10% increase in non-retained new users within the Home feed group compared to Explore
#####Curiosities
- 3.1 What share of users updated their default feed to "For you"?
- 3.2 Do we see a difference in scroll depth from users who configured interests vs those that skipped?
- 3.3 Which modules in the Home feed drive the most engagement (as a share of each tab)?
###Follow along at:
https://www.mediawiki.org/wiki/Wikimedia_Apps/Team/Explore_Feed_Refresh
------