Once you’ve received an alert, your next question is typically: “What happened?” Here are step by step instructions to what you should do to analyze the alerts and figure out the root problem causing the alert.
1. Create a trend report: If you built an alert from scratch, build the same report to visualize how the metric changed over time and specifically during the time of the alert. If you created the alert based on a report, then check out the report.
✏️Note: To view a trend, group the report by time properties.
2. Group report by properties: Group your report by relevant properties such as app version, device models, user segments, or cities. Use these properties to find correlations and locate the main source of change.
3. Hone In Deeper: Once you find a possible problem, such as a problematic country, further analyze the data to delve deeper into the root cause of the alert. Filter the reports by the problematic country and check if the drop began around the same time of the alert, and explore more breakdowns such as operating systems to determine any potential impact factors.
- Play sessions: Watch sessions to gain more insights and validate hypotheses regarding the identified problem area.
- Take Action:
- Based on the insights gained from the analysis, take appropriate actions to address the identified problem.
- Implement changes or optimizations to improve the user experience or resolve any issues identified.
- Continue Measuring: Continuously measure the impact of the actions taken to see if they have helped improve the situation. You can create an alert contrary to what was triggered.
By following these steps, you can effectively analyze and address alerts received through UXCam, ensuring optimal app performance and user satisfaction.
USE CASE - Drop in new feature adoption
You’re alerted via email about a significant drop in adoption for a recent feature launch. You review the report to confirm the sudden drop in adoption.
- By duplicating the original report and breaking it down by various device properties and user properties, you pinpoint that the issue is focused on Android users.
- You further filter reports specifically for Android users to narrow down the problem. You group the filtered reports by device model and identify specific Samsung device experiencing issues with the feature.
- You validate the impact of the specific device model by examining user feedback, error logs, and any other relevant data sources. You communicate the findings and orovide details about the bug and device model optimization issue.
- Once the fix is deployed, you continuously monitor adoption metrics following the implementation of the optimizations. Track the recovery of adoption levels among Android users, particularly for the identified device model, to ensure that the issue has been successfully resolved.