Filtering fundamental data display

Allow users to monitor fundamental survey data with filtering capabilities to directly obtain survey data.

Adding customized data card for better understanding and monitoring

Creating customized data card to gain deep insight on survey

Setting up alert for abrupt data changes

COP will alert users on abrupt data change on workplace tool as they set up.

Understanding users and scenarios

Riley

Jessica

Chenyi

Required
data types

Survey understanding data

Monitoring data

Modeling data

Need

1. Quickly collect results after publishing a survey
2. Analysis data for deeper understanding

1. Quickly collect results after publishing a survey
2. Monitoring the change of the foundation data, presentations rates, and response rate
3. Identify the reason of data trend changes

1. Monitoring the change of the foundation data and other essential data
2. Identify the reason of data trend changes

Pain points

1. Current platform only has survey data for the past day
2. Unable to observe data based on regions

1. Unable to build customized data monitoring
2. Current trend chart only shows UV and submit rate trends

1. Unable to view data based on regions and time
2. The data change alert fully rely on human observation
3. The platform does not display the distribution of survey exposure frequency

User story

Reily launched a new survey on users' sense of control and aims to quickly collect responses to analyze overall submission rates and choice distribution.

Additionally, the goal is to determine whether there are significant differences in submission rates and choice distribution across different regions.

Jessica is monitoring user satisfaction on TikTok content and the trend of popular content. She wants to quickly track the weekly trend of the key satisfaction metric, user like rate.

Chenyi utilized survey samples for model training but has previously encountered anomalies in event reporting, causing a sudden drop in sample size and ultimately impacting the online model.

To prevent this in future, she seeks to implement real-time alerts when a significant decline is detected—specifically, when impressions decrease by over 80% compared to the previous hour or by more than 30% compared to the same hour last week.

1. Understand what data are required from users

After understanding user's need and pain points, I came up with that what data should be displayed on survey data analyzer page and assorted them into groups and sub-groups.

2. Explore layout for better data display

Given the data that are required for users, I began by exploring the layout of the analyzer page. To solve the design problem, I prioritized designing the essential data area and the trend chart area.

In the end, I chose to adopt the third layout which transparently display all core data on individual cards directly without any collapse. The layout is also scalable for adding more customized data card by just having an extra row.

3. The layout of essential data card

Since users monitor eight fundamental data points and customized metrics, I explored various layouts to optimize readability and data processing. I ultimately chose the third data overview design, which organizes information into three categories, making it easier for users to understand.

4. Filter data for deeper understanding

The filter helps users focus on a specific data segment for deeper insights, enabling precise analysis and informed decision-making. I ultimately chose option 2, which maintains the original spacing.

1. Set up alert for sudden data change

Users can set up alerts for sudden data changes, which are automatically sent to their office software - Lark. I ultimately chose option 2, as the number of alert users is typically fewer than five, making it easier to view and manage all information within the alert window.

2. Notify users data change on data card

I also iterated the data card to include notifications for data changes. Explored multiple options for displaying fluctuations and, considering space constraints and the need to present all data, ultimately adopted the narrow card design.