1 of 42

App Strategy

Apps Gratia Appis

2 of 42

Current strategy

Understand our users: Qualitative and quantitative research about our readers

Improve

encyclopedia experience

Engagement and retention in current experiences

Reach new readers in the “Global South”

Collaborative and interactive experiences

EXISTING USERS

NEW MEMBERS

Reach

new readers

Interactivity

with new experiences

Strategic Initiatives

NEW READERS

Based on Foundational Work

Services: Build out services to support new experiences across all channels

3 of 42

In the last year

  • Explore feed
  • Reading lists
  • Deeplinks
  • News notifications
  • Mobile Content Service APIs
  • Wikidata description editing (Android)
  • Places/Nearby
  • Widgets
  • Link previews
  • Two factor log-in
  • Navigation redesigns
  • Visual redesign and brand standardization
  • Fundraising
  • iPad design
  • Find-in-page
  • Features for the visually impaired

… and more...

4 of 42

According to users...

5 of 42

Bottoms up

Both app teams perform a product self-evaluation on 8 dimensions (scores are avg out of 10)

PLATFORM

APPY

STICKY

RESPONSIVE

PRIVATE

POLISHED

ENGINEERED

FEATURED

ACCESSIBLE

iOS

7.25

7.5

6.75

8.25

7.25

8.5

7.5

8

Android

7

3.25

3.75

5.75

7

7.5

6.7

4.7

Additionally the teams voted on priorities they have for Q1 and Q2, winners included:

  • Improve the Explore feed, especially for freshness and customization
  • Iterate on design and finish user experience overhaul on Android
  • Updated and more informative on-boarding
  • Offline content and resource sensitivity for New Readers
  • Continue to work on more testing (automated, QA and user)
  • Reading lists across platforms
  • Analytics

6 of 42

By the numbers

ANDROID PAGEVIEWS

iOS PAGEVIEWS

1.1% of all pageviews (June 2017)

0.5% of all pageviews (June 2017)

7 of 42

By the numbers (cont)

ANDROID 7 DAY RETENTION

iOS 7 DAY RETENTION

8 of 42

Are the apps a “success”?

Great press and partner feedback

Strong momentum

Productive teams

Good research results and user reviews

Integrate design and do it with style

Leading edge of tech modernization

The best experience

Audience is small relative to our web traffic

Audience is stagnant

Specialized development

Quality control is different and difficult

Still controversial in some quarters

Limited editing

(it’s a Wikipedia app, not a wiki)

YES

NO

9 of 42

Opportunities

Wikipedia’s app usage is significantly lower than news app usage.

We believe we offer a great reading experience for high affinity readers. But they need to know about the app.

10 of 42

Are readers going elsewhere?

Three highest ranking Wikipedia focused apps on Android and iOS share a few core features, which our apps also include, although in some cases the “completeness” of our feature is comparatively poor.

APP

DOWN-

LOADS

OFFLINE

NEARBY

CUSTOM UX

LISTS

1M-5M

x

x

x

x

500K-1M

x

x

x

x

500K-1M

x

x

APP

RANK

OFFLINE

NEARBY

CUSTOM UX

LISTS

~200

x

x

x

x

~400

x

x

x

x

~700

x

x

x

x

Wikidata Properties

11 of 42

Are we capturing new installers?

We are near benchmarks for users who do come to install, although there is potential in further localization and market message testing:

12 of 42

Awareness

Without Google to send initial pageview, users must discover the app on their own. When asked, during the 2014 consultation about where the foundation should focus for the future, Mobile Apps were the number one request among anonymous users, even though, at the time, the apps existed:

That hasn’t changed:

13 of 42

Awareness channels

CHANNEL

PROS

CONS

EFFECTIVE?

Wikipedia

We are sitting on ~1B monthly users who already use our product...

Community concerns likely. Need to be clear and reasoned about who the apps are for and what value they provide users and the movement.

Yes

Stores

Relatively cheap

Requires BD. Sometimes require technical effort. Limited duration.

Yes, but very short term

Preloads

Leverage brand for free placement

Shrinking market, requires BD, retention unknown, increasingly savvy users do not use preloaded apps - test with Ironsource going now

TBD

Press

Record of success, impact on iOS

Requires comms/PR firm. Requires product focus on new features. Not impactful on Android.

Mixed, but very short term

Platform integration

Provides a built in mechanism for discovery.

Beyond what exists, requires significant BD effort and value for partner.

TBD

14 of 42

Who are

apps for?

For high affinity readers

For news readers

For new readers

For new contributors

For existing editors?

For our technical ecosystem

For the Foundation

For the way we work

For the mission, of course

Photo by Mauro Mora via Unsplash (CC0)

15 of 42

Who should we build our apps for?

People who love Wikipedia, even if they don’t know it yet. They are curious, want to learn and see the app as a way to understand their world, rabbit hole on their interests, and pass the time while mobile.

Photo by Raw Pixel via Unsplash (CC0)

16 of 42

What do we do for them?

Provide real value by enriching their daily lives with knowledge, making learning from free sources a habit woven into their days. The people who love to learn and regularly read Wikipedia deserve an app that loves them back. An app always there with an interesting suggestion or an in depth dive on the latest news. They want to hear from us when someone important dies. They want to wake up and find out what the rest of the world is curious about. The apps can do that.

17 of 42

Apps are for lovers

Photo by Luke Porter via Unsplash (CC0)

18 of 42

I’ve always loved Wikipedia, since it started popping up in search results back in the day. I didn’t know they had an app, but I saw a banner on their site when I was browsing some articles about the crazy events happening in our world.

I love the app, I check the feed on my commute, and particularly love the mix of news-y stuff and really random stuff I’d never have searched for on my own. The personal recommendations are meh, though. I really like how I can see something I want to read later, and add it my reading list to pick up on my tablet at home. I have like a hundred articles saved up!

From using the app, I also feel like I understand wikipedia better than when I just got to it from Google searches. In this time of fake news, and where everyone tracks what you read to sell you stuff, I really appreciate having a free source of non-commercial information at my fingertips.

19 of 42

How many people love Wikipedia?

We don’t know. But as a first estimate, we started by looking at devices on the mobile web site which already visited yesterday.

English Wikipedia (en.m.wikipedia.org) is about 8-10 million per day.

For Spanish, 1-2 million.

Compare to e.g. the 1.1 million daily active users of the Android app.

DAILY ACTIVE USERS

20 of 42

Audience awareness impacts

So each day millions of devices visit wikis they also visited yesterday. If we could target users returning within the next day, how would the apps grow?

Daily Install Funnel

Additional users on Day 1

21 of 42

Features for lovers

Q1

  • Reading lists that sync across devices
  • Feed improvements, particularly to improve customization and content freshness
  • Reading themes, including dark mode

Q2

  • Navigation and onboarding improvements
  • Queueing and rabbit holing experience
  • Better recommendations

H2

  • Notifications

Many of the features developed in the last year will be ported across platforms and

studied, refined and iterated.

22 of 42

Apps are for news readers

Photo by Toa Heftibar via Unsplash (CC0)

23 of 42

2nd stop on the 2nd screen

Media and events are significant motivators of use

In the Knight study of news consumption, we are not the source of news, we are the second stop.

The need to fact check everything, means having Wikipedia on hand (ie. on a tablet or phone) while consuming other media.

This is exactly why mobile is called the “second screen”.

Filling this need is both mission aligned, timely and potentially very impactful. But, we have a lot to learn here.

24 of 42

Features for news readers

Q1

  • Trending-by-edits research and potential iOS integration
  • Research on feed use and relevance of trending and new content
  • Update Android news card experience

Q2

  • Expand use of news portals and research ways to offer more tailored news by geography or interest
  • Continue research and prototyping of trending, including trending searches

H2

  • Scale up and expand trending and news notifications

25 of 42

How do we reach them?

Target announcements on wiki based on specific target audiences, and their likely affinity for app usage based on specific behaviors.

Examples of targets and messaging around specific features based on observed usage:

  • highlight Trending and Notifications features for users on currently trending or news articles
  • notices for users returning within 1 day, or reading many articles in a single session on mobile

Consider working with non-en language communities with high app/smartphone usage and some endemic interest in supporting apps. For example JA, KO, ZH, and some western european languages.

Learn more about what’s happening today in your world.

Our app makes it easy.

Download the Wikipedia app today.

26 of 42

Apps are for new readers

Photo by Ntsika Lain via Wikimedia Commons (CC by-SA 3.0)

27 of 42

Who could get through grad school without Wikipedia?! But I’m a student, and I don’t have any money to spare on data plans. So, I planned to download all the articles about this semester’s classes, while I’m connected to the school wifi. I was saving articles to my SD card by printing them to PDFs. It worked okay, but they were a pain to download one at a time, and hard to search through.

During one of my download marathons, a banner popped up for the Wikipedia app, and it's saved article libraries. Once I saw you could download entire collections of articles about a topic, I got the app downloaded a bunch of stuff for school. Now I use the app for homework or just learning more without using any data, and when I need to look up something I haven’t downloaded I can use the same app while I’m on my school WIFI.

Photo by ilaya raja via Unsplash (CC0)

28 of 42

New readers features

Q1

  • Support for offline compilations on Android (Kiwix file support)
  • Better local data management

Q2

  • Expand data consumption options
  • Improve visibility around bandwidth and resource usage

The apps teams also are working with New Readers to determine the best way to partner with Kiwix, the incumbent non-commercial offline Wikipedia app.

29 of 42

How do we reach them?

Target announcements on wiki based on specific workflows which are likely to be usds by new readers. Use context information like country and connection speed to identify users who might benefit from the apps.

Examples of targets and messaging around specific features based on tasks include:

  • Tell mobile users saving a lot of PDFs about offline support in the apps
  • Tell users switching frequently between languages about the apps multilingual support

Work with New Readers program and communities with zero support and community interest, to raise awareness of offline capabilities of the apps.

Saving articles to read them later?

Our app makes it easy to read Wikipedia offline.

Download the Wikipedia app today.

30 of 42

Apps are for new contributors

Photo by Jason Ortego via Unsplash (CC0)

31 of 42

I always read Wikipedia’s pages on my favorite shows but I didn’t really care about editing it. When you click on editing on your computer it looks like something I’d use for school work. I also used the app, since I like to read on my tablet, and noticed you could edit the little description under the title. That seemed like a lot less work, and many of the descriptions were missing in Korean (though I often read the English articles because there are so many more of them about everything). After I realized the mobile editing was a lot simpler, I started adding citations. Now, when I see the citation needed banner, I’ll open Google and Wikipedia open side by side, and find the link to add. It’s easy, and I like helping other people learn about the shows I love.

Kyong, 22 yo Casual Editor from the Republic of Korea

Photo by Jody Hong via Unsplash (CC0)

32 of 42

Features for new contributors

In late 2016, Jon Katz and the Android team engaged in an on-wiki consultation about over a dozen potential mobile contribution features. The upcoming annual plan includes Program 3, which will deliver 1-2 additional contribution features in the apps. The community favorites included:

Wikidata Properties

Wikilabels for ORES

Add Nearby Images

Thank the Editors

33 of 42

Apps are for mobile editors?

Photo by Joshua Newton via Unsplash (CC0)

34 of 42

Most platforms provide full editing functionality on mobile via apps. Apps provide the best user experience and performance for complex tasks like WYSIWYG editing on mobile devices:

35 of 42

Apps are for our tech ecosystem

Photo by Victorgrigas via Wikimedia Commons

36 of 42

Apps enforce a separation between clients and data and provide an in-house customer for a platform agnostic and interface neutral stack:

Apps also provide a place for people with mobile app expertise and interest to contribute. Both apps regularly receive contributions from multiple volunteers in each major release.

All the knowledge for all the people...

MediaWiki

m.

Apps

All the knowledge for all the people...

Services (build once)

Web

Apps

m.

MediaWiki

Web

The next mobiles

37 of 42

Apps are for the Foundation

38 of 42

For the Foundation

  • Tell audiences that Wikipedia is a modern, relevant project
  • Burnish our brand in the tech and media industries
  • Provide space for innovation without as much risk to the core
  • Maintain a positive relationship channel with two important platform owners
  • Serve and connect with readers directly, without search engine mediation
  • Higher per-donation fundraising results than mobile web
  • Help to attract new talent, particularly in design staff and user oriented developers
  • Provide a way to turn high frequency users into brand advocates

39 of 42

Apps are for the way we work

40 of 42

The apps have been delivering significant and well received improvements over the last 18 months. That’s what we care about. But we’ve also worked to level up the way we build software at the foundation, with a particular commitment to listening to our users, and working to integrate qualitative and quantitative data into our work.

  • Iterative agile process
  • Surveys and studies of existing users
  • Personas and user stories
  • Chore wheel keeps us listening
  • Evaluative user testing of new features
  • New app analytics project

Clearly identifying our audience, trying to understand them, working across product, design and engineering to give users the software they want.

41 of 42

Apps are for the mission

Photo by Niccolo Caranti via Wikimedia Commons CC BY-SA 4.0

42 of 42

1. Why build apps?

  • For readers who love Wikipedia
  • For new contributors to Wikimedia
  • For Wikipedia Editors who want the nicest mobile editing experience?

2. But, why can apps do that?

  • Because apps provide unique technological capabilities, audiences and user expectations.

3. But, why does that matter for the mission?

  • If Readers have a more intimate Wikipedia always at hand, it helps to keep Wikipedia relevant, making them part of the movement, and embedding the mission deeply into their daily lives.
  • If we can design and socialize a mobile contribution system that respects our curation norms, we can tap into even more knowledge from more people.
  • If we can make editing and curating awesome in the apps we can provide our editors with the best possible tools for building, curating and administering the encyclopedia on mobile.

I only got 3 why’s deep.. ¯\_(ツ)_/¯