Jisort! DSO Software Prototyping Update

Following from Tamasha Hack back in April, the DSO Software Group pursued a friendly, interactive tutorial app to give new smartphone users the confidence and agency to solve problems with their smartphones on their own.

The result is now ready to be put into their hands! Introducing...


Jisort! is an informational prototype app that and offers fun and visually interesting ways to increase basic smartphone awareness and skills, diagnose problems, and encourage exploration.

Ji: a Sheng term added to Swahili words to imply "do it yourself"
Sort: to solve a problem or deal with something pending

Jisort = Ji+sort = "Sort yourself out"

Insight gathered throughout the project suggested new smartphone users' problems and curiosities are focused on a few topics, like data usage, storage space management, battery life, literacy, and accounts.

The app is available at mzl.la/jisort and see other release builds on the github page. This version works on android 4.2 and above. Keep in mind that this is a prototype Android app, designed to test certain concepts about learning digital skills through smartphones. We are actively collecting feedback from the community and study participants on its efficacy.

What's in the box?


The majority of Jisort is comprised of tutorials aimed at performing tasks or answering questions about smartphone basics.

  • Wi-Fi ni Noma (Wi-Fi is Awesome): All about Wi-Fi and its data- (and therefore money-) saving benefits.
  • Connecting to Wi-Fi: Instructions and simple troubleshoot for connecting to Wi-Fi.
  • Data Consumption: Understanding how to track data usage on your phone.
  • Using Data Wisely: Tips for managing your data consumption.
  • Downloading New Apps: All about the Play Store, what it provides, and how to use it.
  • Freeing up Storage: How to think about and preserve space on your phone.
  • Using Airplane Mode: Why there's an airplane icon, and how it can help you manage the wireless services on your phone.
  • Accounts and Passwords: The importance of account awareness in the era of apps, and how to create strong, memorable passwords for them.

Each tutorial is filled with emojis and icons to provide small reminders and necessary context for some tricky icons or concepts. And, to avoiding large pieces of text, each tutorial step has an image or animated gif to further demonstrate concepts or illustrate navigation through UI. Our hope is that animated gifs make instructions easier to grasp or at least more visually stimulating.

With advice from participants and the community, most of Jisort's content is written in English, since Swahili either lacks the vocabulary, or requires to many words to describe common smartphone concepts.


Some tutorials are linked to specific Android events that Jisort can detect, like connecting to Wi-Fi, or having low storage space. When these events occur, Jisort uses a native Android notification to tell the user that they can learn more, offering the user opportunities to better understand and manage their phone.

Icon Quiz

To contrast Jisort's instructional content and to assist in English and digital literacy, the Icon Quiz is specifically designed as an interactive learning opportunity. It has two parts:

First, the Find an icon screen lets users learn about icons they may have seen on their phone, or discover new ones they will encounter inevitably.

Then, when they feel ready, clicking the PLAY THE ICON QUIZ button will let users test their knowledge through a small quiz.

Space Challenges

Jisort weighs in at about 10MB as an APK, and unpacks to become about 14MB. Compare this to Facebook's 36MB download and 150MB+ operating size, and Whatsapp's 13MB download and 50MB operating size. While testing in the field with DSO participants', restrictions on storage space still made the app difficult to install.

Many DSO participants' smartphones run Android 4.4.4 and below, which use partitions to separate apps and other system data from media, downloads, and other user-generated content. Consequently, phones with 1 or 2GB of internal storage may have only 100MB or 200MB available for apps. With Messenger, Whatsapp, a couple games and some bundled apps, there is little room for much else. Users are forced to decide between what they want on their phone, and what they need.

Some study participants navigate this problem somewhat gracefully with the help of an SD card and APKs. Already, many people are forced to find ways to get apps without using the Play Store, so it's natural to keep a library of APKs to unpack and install on-demand. However, many users are plagued by this problem, and quickly abandon attempts to download the apps they want.

(Facebook Lite already has the right idea here, available at ~1MB, although it lacks much of the modern styling many Facebook users are to.)

The Cutting Room Floor

Jisort's design was very aspirational, and, as with most aspirational projects, features were cut that we would love an opportunity to explore more.

Sheng and Swahili

Originally, Jisort was intended to contain Sheng where possible. However, the final build contains only one or two references.

Content creation that incorporates Sheng was difficult from the outset. While community members were quick to try, concepts were already difficult to explain, and would be more complex with the forced introduction of Sheng.

Content was written in English first (because it's the common language of the development team), but perhaps a native Kenyan content writer would have been able to approach content creation differently with a mindset toward simplified, Sheng-first language.

However, with respect to written language, there is also a question of formality. English is something of a standard for education and business—even popular Kenyan banks apps support English at launch. Anecdotal evidence from participant interviews suggest that English is preferred and respected, but the roots of that sentiment are unknown.

This language divide is very testable throughout the rest of the Digital Skills Observatory. Upcoming interviews and the next software prototype should incorporate methods to understand

  • which type of language people are more likely to comprehend on the topic of technology, and
  • in which language they prefer to learn and be addressed.


To further explore the edges of language literacy and and smartphone technology, one of the early proposed ideas included a text-to-speech mechanism so that paragraphs and icons could be spoken to the user when clicked.

Unfortunately, the Swahili text-to-speech was not reliable enough to explore this option. However, English should have been considered as a viable option, since simple English is well understood by most participants, regardless of their literacy level, especially with a Kenyan accent.

This feature is also testable, even without building another piece of software. Android has a powerful text-to-speech engine that Jisort would have used, which could be explored alongside participants as they search the web for favourite music and apps or and attempt to explore the settings of their phone.

Floating Action Button

As we wanted to test new ways to onboard new Android users, Jisort had a Floating Action Button (FAB)—like Facebook's chat heads—throughout most of its development.

The FAB was intended to make finding Jisort easy, so that people don't need to go searching for it when they run into trouble. Additionally, Jisort's FAB was designed to light up when it detected a problem or a potential learning opportunity.

However, with respect to apps, there have not been many phone navigation problems observed among participants. Coupled with difficult implementation (with lots of bugs and not enough time to iron them out into an acceptable UX), this feature was removed in favour of Notifications discussed above—although even the learning opportunities in that feature leave much to be desired.

There is still an opportunity to explore first-time-use experiences for Android using this or a similar feature.


One of the more intriguing features that had to be removed early on due to time and resource constraints was the Screenshot help feature.

If implemented, users would have been able to take a screenshot of a problem they encountered, and use Jisort to send it to a digital skills expert (likely a member of the DSO community). For users that could otherwise not explain their problem in detail, this feature would have provided learning opportunities for the study participants and the researchers. (When a user has a problem, there is likely insight to gain!)

For example, when an app crashes with a cryptic error message, or an app can't be installed due to mysterious space constraints, users could simply send a picture of what they see through Jisort and receive a response like, "How often does this error occur?" or "Did you receive this app from a friend over Flash Share?" or maybe "Go to the Storage Settings panel, and show me how much storage space is used."

Unfortunately, with further consideration, this feature would have been fairly data-heavy. However, integration with Whatsapp or Facebook Messenger would have been fascinated to explore, especially where Whatsapp groups or Facebook pages could have been used to talk about problems among other users, providing community learning opportunities.

Deployment and Follow-up

Over the course of the following weeks, DDD's researchers are installing Jisort on willing participants' phones in study's treatment group.

During its use, Jisort will collect analytics about what people choose to click on and read within the app. Since mobile data and storage space are such valuable resources, these analytics are stored in a size-restricted file called .jisort_analytics.txt in the public Download directory. Researchers can use a proverbial Konami Code to invoke the Android share feature and collect the analytics file:

  • Open the first screen of the Icon Quiz
  • Click on the Wi-Fi icon 10 times
  • Long-press the Wi-Fi icon

Upcoming interviews and intervention workshops will be important learning moments for DSO's research team, as participants will be able to explain their joy and frustration and provide feedback for Jisort, especially in contrast to the workshops they've attended. If participants keep Jisort on their phones and use it regularly, it will reveal more about how people prefer to learn about their smartphone.

The DSO team has learned a tremendous amount from study participants and the design process of Jisort. Looking ahead at a second prototype, the development of Jisort left behind many ideas still worth exploring. However, it is not clear that another app is worth building, or if other smartphone-centric ideas should be pursued.

Do you have an idea? Reach out to dso@mozillafoundation.org.

The code for Jisort is available on github, and we encourage it to be forked and updated to pursue more exploration in this space. Don't hesitate to reach out here, on github or at dso@mozillafoundation.org.

None of this work would have been possible without the talent, patience, and effort of the DSO Software Group contributors, including Festus Langat, Dobbz Mkolwe, Brian Mwadime, Walter (Loso) Obadha, Mike Otieno, Luke Pacholski, Laura de Reynal, Bobby Richter, Chad Sansing, Alex Wafula, Stephen Wanjau, Faith Zuma, and the rest of the DSO Community.

View Comments