A rapid approach to understanding the user experience of encountering errors: Feedback analysis

It is obvious that when users encounter an error, a bad experience is provided for them, no matter how small the error is. Therefore, we always try to identify user errors and reduce the probability of their occurrence. In this story, I want to explain why the installation error in the Bazaar app store became one of the main concerns of the development team and how we used feedback analysis to understand its dimensions in a day.

Introduction

‌Bazaar is an app store with more than 40 million users, which complicates comprehensive user research. After numerous usability testing projects and in-depth interviews, we reached 3 main journeys of our users in installing the application.

With a quick look at each journey map, we realized that the most painful part of the user experience was exactly the moment of downloading and installing the program. Where almost the most vital purpose of the App Store is there. The image below shows the shared feeling of different user groups at the end of each journey.

So we decided to investigate this issue deeply. We wanted to know why these errors occur. How to reduce them or make it easier for users to fix errors at least?

Method

In choosing the research method, we had to consider two points:

1- Errors are sudden and unexpected events in the way of user experience. Obviously, the user does not remember them completely in the interview or survey.

2- If I had chosen test or laboratory conditions to investigate this issue, there would be a high possibility that the user’s accuracy in dealing with errors would be very different from real conditions.

So, for the first research, I needed data to understand how users deal with these errors in real situations. What do they think of them, and how do they try to solve it? In this case, I decided to review the support calls of the last few weeks. The operation people had well-tagged the calls by topic, which made my job very easy because I could filter to hear only the calls that were related to the users’ problems in installing the app.

Execution

I listened to about 40 calls about installation issues and reached saturation. I documented each issue on the board. Then I did a thematic analysis. Finally, I came across several categories of basic user problems at this stage, making it more difficult for users to fix the error.

Here I explain some of the insightful tips from calls:

1- what we didn’t think about was kids. Through this research, I found that encountering errors by kids is a completely different story. When kids have problems downloading the game:

  • Most don’t want to solve this problem with the help of their parents.
  • Some of them cannot read the error code and description.
  • Most of them could not explain their problems clearly to others.
  • Sometimes, they are embarrassed to talk on the phone
  • Those who are too young do not understand the support team’s instructions.
  • But, they insisted on solving the problem to become able to download games and play.

2-Instead of reporting the error, we used to report the error code. Some users did not see the error code at all.

3- They had to report their Android and Bazaar versions to be guided when they called support. Users did not know how to find this information.

4- Sometimes, due to lack of free space, users face problems in installation and need to manage the space of their phones, but because they think that apps occupy more space than video and music files, they can hardly free up space for new apps

5- Multiple errors: Sometimes, the party’s conditions are subject to several errors. We will show them one by one. For example, at first, it gets error 110; after a few calls, when the problem is solved, it gets error 4. This makes the user tired and nervous and prolongs the error-fixing process.

Solution Part

In several workshop sessions, the team (the product manager, UX people, and engineers) checked the users’ problems in the installation and brainstormed some solutions. We have chosen and planned to develop several time-consuming solutions with the aim of fundamentally reducing the possibility of user error.

But due to the sensitivity of the “download” issue for the App Store, we considered 3 short-term solutions to reduce the difficulty of this part of the user’s journey:

1- A service where users can send us the information needed to fix the error with their permission and with one click.

2- Display the error description instead of the error code when we inform the user about the error for the first time

3- Rewriting the explanations of errors that were sometimes vague.