Session insights have transformed how we detect user errors and prioritize fixes across regionsGlassbox now has GIA as an AI integration that provides analysis very quickly without needing to watch the entire session. Based on previous results, the analysis is accurate 90 to 95 percent of the time regarding possible reasons why a user faced an issue. Based on that analysis, we can make a report and get things done to prevent errors. The best features that Glassbox offers are reporting and the funnel. These two use cases help us most as we can determine the expected count for user behavior and compare it to earlier weeks and days from the same period. In the report section, we can identify specific alerts that are being triggered on a regular basis so we can capture and analyze them and fix them in the longer term. The funnel feature helps us check user impact and identify which countries or specific areas the website is failing in. With this feature, if a customer is facing a cart page error, we can check how many users were impacted at that specific time and which regions those users are experiencing errors. We can also check what products those customers are trying to add. Based on that information, we can create a report and check comparisons. When we share it with the business team, they can determine if the issue is widespread and work on it with higher priority, or if it has lower impact with fewer users affected, indicating an intermittent error that can be worked on with lower priority. Glassbox has had a positive impact on my organization through all the features it provides. As AI is advancing and its adaptability within the app itself integrates many use cases, the reports, funnels, dashboards, GIA features, and integrations that Glassbox provides help us address issue behavior and work on resolutions. Additionally, it creates reports that we can share with other teams so they can determine if the impact is wider. To prevent errors quickly, Glassbox captures reproducible customer errors. If an error is not reproducible, Glassbox does not capture it and can mark it as a false alert. This is how we prevent false alerts and capture only genuine alerts.
