Wikimedia Apps/Offline support/V1 User research
|This page is currently a draft.|
More information and discussion about changes to this draft may be on the talk page.
- 1 Overview
- 2 Research findings
- 2.1 Research reports
- 2.2 High-level findings from both studies
- 2.2.1 1. Utility is conditional to pack size & specific subjects
- 2.2.2 2. Marketing/User-education was understood but not compelling until users were offline
- 2.2.3 3. More clarity/details about storage usage was desired by some users
- 2.2.4 4. Usability improvements identified to improve the searching and downloading flow
- 2.2.5 5. Users wanted to view more relevant article packs upfront
- 2.2.6 6. Concerns about wasted usage of mobile data during downloading
- 2.2.7 7. Many users did not understand the on/offline status affected whether they were viewing articles from the Offline Library.
- 2.2.8 8. Most people used the 'Main page' button in the downloaded pack details as the entry point to using article packs
- 2.3 Recommendations & Next steps
In September 2017, a V1 prototype of “Offline Library” created on the Wikipedia Android Alpha app was tested on with users in two studies. The first study with the primary target audience of New Readers was conducted by Hureo, a third-party user research agency, with participants in Pune, India. The goals of the study were to:
- Discover whether participants find the "Offline Library" feature useful and if they are successful in fulfilling their offline information needs.
- Identify usability improvements.
A second unmoderated study was then conducted via Usertesting.com with a secondary group of "Active readers", with the same goals in mind as above, as well as to compare differences in the utility and usability of the feature between the two groups.
- Report for Android app Offline Library V1 in India (PDF via Hureo)
- Offline Library v1 - 'Active readers' user testing results (Google Presentation)
High-level findings from both studies
1. Utility is conditional to pack size & specific subjects
The concept was well received by participants, but practically their usage was quite conditional to being able to download packs for specific subject areas rather than our hypothesis that users would want the ‘whole’ of Wikipedia.
KEY RECOMMENDATION: Make smaller, more specialized content topics for article packs. More details about the types of content users wanted to access can be found in the respective reports, but they ranged from articles about science, mathematics and other school curricula subjects, to more constantly curated collections dedicated to trivia and "current events".
2. Marketing/User-education was understood but not compelling until users were offline
Understanding of the feature was generally good when users reviewed onboarding and promotional information, but the term 'packs' was still somewhat unclear until users started using the Offline Library
KEY RECOMMENDATION: Reconsider the term 'article packs', and make promotion of the feature more contextually-relevant.
3. More clarity/details about storage usage was desired by some users
Some users were unclear as to whether the amount of storage used by the Offline Library was internal or external, more specificity on file-size units, and generally desired more options for obtaining, storing and sharing packs.
KEY RECOMMENDATION: Show both the internal and external storage space being used by article packs, and provide an option (when available) for users to choose an internal/external download destination.
4. Usability improvements identified to improve the searching and downloading flow
The call to action to "Add to my Offline Library" was not clear to many users as being the same action to search through available offline packs for download.
KEY RECOMMENDATION: Make the “Add to offline library” action as prominent before and after a user has downloaded an article pack.
5. Users wanted to view more relevant article packs upfront
Whilst searching for article packs was easy, some users wanted to view more relevant article packs and see more information upfront about each pack in order to help make the decision as to whether they should download it.
KEY RECOMMENDATION: Add a clearer visual indication in Offline library list for already downloaded packs and surface relevant article packs more clearly by filtering to relevant languages by default.
6. Concerns about wasted usage of mobile data during downloading
More error-prevention desired by some users that they were not unduly using data.
KEY RECOMMENDATION: Prevent data-usage waste by showing messages preventing/warning against downloading when storage is insufficient or downloading using mobile data instead of WiFi.
7. Many users did not understand the on/offline status affected whether they were viewing articles from the Offline Library.
Many users did not realize that they would not be using the version of an article from the article pack if they remained online, but rather the app would continue to load the most recent version of an article. Similarly, users were confused that they had to be offline in order to restrict searches only to Offline Library packs.
KEY RECOMMENDATION: Provide clearer details on a users' on/offline status and allow users to set data-savings preferences to keep viewing the downloaded version of articles found in their Offline Library whilst online.
The majority of users 'accessed' their Offline Library content this way rather than knowing that they could search for articles whilst offline or by entering the main page for articles packs whilst offline via the 'Explore feed'.
KEY RECOMMENDATION: Update and/or standardize the format of the "Main page" across all article packs to show clear ‘introductory’ text and high-level content categories as a starting point for exploration (similar to the main page of the WikiMed pack).
Recommendations & Next steps
The full set of recommendations is detailed in the document Offline Library v1 - Recommendations & Next Steps (Google Presentation).
In addition, the Phabricator task comment  tracks all updates made to the Offline Library feature based on the feedback from this round of testing.