New Readers/Retention

The New Readers program seeks to expand the reach of Wikipedia in emerging markets, where usage is lagging behind internet penetration. The overall goal is to drive growth of Wikipedia in emerging markets through increasing readership, particularly on mobile devices. More engaged readers should lead to more editors and therefore more content where Wikipedia currently has gaps.

Defining the challenge

Mobile internet penetration is expected to grow from 28% to 45% in developing markets from 2014 to 2020, an increase of 700 million potential Wikipedia readers and editors. Research in 2016 indicated low awareness of Wikipedia for both online and offline readers in key markets in emerging countries.

Country code	Country name	% of people who report having heard of Wikipedia (internet and non internet users)	% of internet users who report having heard of Wikipedia
NG	Nigeria	23%	27%
IN	India	22%	33%
MX	Mexico	45%	51%

The Wikimedia Foundation New Readers team has since made subsequent pilot interventions in marketing, product, partnerships, and grants. As a result of these interventions, the Product team is ready to create prototypes with a series of product changes that will engage and delight new readers of Wikipedia. This product exploration is aimed for users who will be using mobile platforms for their primary online presence, and in many cases, as the only way they go online.

Product goal

Deliver compelling experiences to Wikipedia readers in emerging markets, driving increased adoption and retention of Wikipedia.

Implementation Approach

The New Readers Product team will work with a design agency based in emerging markets, with experience in designing products for the types of user personas identified in the 2016 research. These prototypes will be tested in India. The team will work with a design agency based in India to:

Re-evaluate and validate the relevance of our research findings from the 2016 user research
Create 3 - 5 concepts and high fidelity prototypes
Test the prototypes in the field with potential Wikipedia users

Deliverables

Functional prototypes of 3-5 product concepts for Wikipedia, to be tested in India by the Wikimedia Foundation.

Timeline

December 2018: Agency recruitment
January 2019: Start of contract
January - February 2019: Concept development
March - June 2019: Development of concepts, prototypes and usability testing
June 2019: Delivery of final recommendations

The Wikimedia Foundation will take winning ideas to larger-scale testing following delivery of the prototypes.

2019 Product Team

Wikimedia Foundation

Angela Muigai - Senior Product Manager

Anne Gomez - Senior Program Manager

Nirzar Pangarkar - Design Director

Runa Bhattacharjee - Senior Engineering Manager

External Partner

To support our design work, we will select a design agency in India that has experience designing and building products for new online users, particularly on mobile. The agency we select will meet the following criteria:

Competency & Quality
1. Have a portfolio of high quality creative work 2. Knowledgeable about local user requirements, evolving platform adoption patterns, and regional industry practices
Project Fit
1. Coherent details of work plans on the proposal 2. Ready to deliver on budget and within schedule 3. Open to work with international stakeholders
Mission Fit
1. Aware of the Wikimedia movement 2. Aligned with free knowledge and an open culture of sharing 3. (Bonus) aware of the value of community participation & stakeholding

Product Experiments

With the goal to deliver compelling experiences to Wikipedia readers in emerging markets, product concept explorations for the New Readers Program began in February 2019. The Wikimedia Foundation hired a design agency, Uncommon, who have experience designing products for Indian users. Their task was to develop 3 - 5 concepts and high fidelity prototypes for testing with current and potential Wikipedia users. The prototypes will be tested in India, and will be geared towards a mobile web experience.

Experiment Design Priorities

The WMF and Uncommon teams held a design sprint in mid February 2019, to determine what the prototyping design direction would be. It was decided that the prototypes would:

1. Allow users to explore content and related articles easily in the way they use their language day to day

2. Lead Wikipedia's browsing experience to feel less long form and more “snackable” (the ability for a user to get a comprehensive understanding of a piece of meaningful content in a short summary)

3. Make Wikipedia delightful to browse

Participant/User testing approach

The Uncommon team will have a sprint, where the team will propose and prioritize features to be tested. 5 - 8 participants would then test an interactive prototype with the features agreed within a 7-10 day period. Their feedback would then be synthesized and improved upon in the coming week, and the process repeated again, until the concepts are refined enough to take to engineering to build prototypes that can go live.

Wikipedia Reader Journey for Sukhwinder

The participant selection will be based off of the target persona of the experiment, Sukhwinder. Particularly focused on the engagement and retention sections of the reader journey on Wikipedia.

Monthly updates

June 2019

The reading research experiment, with cricket as the topic of choice, kicked off on 6th of June.

May 2019

May was spent putting the prototype together by pulling content from Wikipedia both manually and automatically. A prototype landing site was created, at cricketinfo.io where readers could view content in either English, Hindi or Tamil.

April 2019

After interviewing potential users in February and March, factors within which a reading experiment could be run to get meaningful results for product interventions became more clear.

The reading concepts would be focused towards young internet users in India who are uncomfortable with English
These internet users prefer rich media and short snippets of information compared to reading long form content
The reading concepts/prototypes will only use the content we currently have available on Wikipedia and Commons
The concepts being highly experimental would not necessarily be what would end up on Wikipedia. It is an opportunity to learn first from potential users, then decide what could go into Wikipedia
The content would be in English, Hindi and Tamil
The content not available in Hindi or Tamil would be translated for the experiment
The reader would first see a "snackable" experience, and then have the option to "read more" or "continue reading" to see all the content that is usually in that article on Wikipedia
The content would provide attribution to Wikipedia and Commons for content and images
The experiments would not redirect users from Wikipedia
The content in the experiment would not be editable to begin with as the experiment is focused on reading for now
The content would be a subset of Wikipedia content on the topic of the Cricket World Cup to be relevant to what newsworthy information would be sought at a certain time periods. As a result, the prototype would start to be usable at the beginning of June 2019.

Within these constraints, the learning objectives were then refined as follows:

Learn if snackable content is a driver of increased readership during the duration of the experiment
Learn if snackable content has stickiness that drives users to return to the experience
Learn if language switching when content is available is an option users would choose
Learn which features of the snackable experience users engage with the most.

Target reader:

The profile of our target readers who would land on the prototype would ideally be:

Age 18 - 25 with 75% men and 25% women
Have an android smartphone
Frequently search for information online
Not comfortable reading high school level English
Can read Hindi or Tamil
Has some awareness of Wikipedia but not necessarily how it works
Study university in a language medium other than English
Use products such as tiktok, YouTube, Google Search, WhatsApp

March 2019

March was focused on continuing to seek input from users who match our user persona. The input took into account their feedback on reading concepts

User testing was done on 27th and 29th March 2019 in Indore, MP & Varanasi, UP
Features tested in Indore and Varanasi included:
- Voice search within an article
- Stories within an article (an image heavy, 4 screen scrollable experience, with short sentence captions at the bottom of each image)
- Page previews for mobile web
User tests were held in Bangalore with 7 participants on March 13th 2019.

The users in Indore and Varanasi were a good match to our user persona. They were studying university in Hindi medium, and even though they knew English, they preferred to read or consume content in their language. It was easy for them to identify how to switch languages. They indicated a preference for the rich media shown together with the content, and that they could share it.

February 2019

A subset of the WMF audience team travelled to India to meet with Uncommon, the design agency, for a design sprint. The sprint would be a way to clarify what goals were intended for the reading experience and then ideate and come up with concepts. The persona that was focused on was Sukhwinder, a reader who is not comfortable with English and uses Wikipedia occasionally.

A user test was run at the end of that week with 7 participants
The concepts that were tested at this sprint included:
- The ability to identify and switch to the preferred language for reading
- A summarized short version of a Wikipedia article with quick facts
- The option to read long form content would be accessible through a "Continue reading" button
- A video would be shown right after the user reads the introduction paragraph of an article

January 2019

In late January 2019, Uncommon was selected as the design agency WMF would work with for design concepts, after interviewing 9 other design agencies in India.