Project Duration

February to March, 2024 (6 weeks)

My Role

Sole designer, responsible for UX and UI design

Team

1 Product Manager, 1 Frontend Engineer, 1 Backend Engineer, 1 QA Engineer, 1 Technical Writer

Overview

Who is Treasure Data

Treasure Data has the leading customer data platform (CDP) product, which were mainly used by marketers to create personalized marketing experience for their customers.

What is Segment Insights

Segmentation is the process of dividing a large group of potential customers into smaller, more manageable groups based on certain characteristics so that businesses can target their marketing efforts more effectively. A segment is a group of customers.

Segment editor is the tool to create segment. Segment insights is part of segment editor that provide insights about customers when marketers create segments.

Segment insights feature has two components.

First, user configure insights charts.

Second, user see charts while creating segments in segment editor which provide them insights about their customers.

Project Goal

When I took over the Segment Insights project, Milestone 1 of this feature had already been shipped.

After collecting feedback from customers, we prioritized the following as the project goal for Milestone 2:

Adding more chart types to enable marketers to view additional insights while creating segments.

Constraints

Limited space on segment editor UI to display charts. Panel cannot be expanded.

Process

1. Understand the Problem (Ask)

I start off by discussing customer’s ask with the Product Manager.

Customers asked for three types of additional charts:

Cross tabulation chart
Ranking chart
Count distinct chart

For cross tabulation chart, the use case customer provided is they would like to see data visualization for age * gender distribution.

2. Research the New Chart Types

There are several types of cross-tabulation charts, including Stacked Column Charts, Clustered Column Charts, Heatmaps, Tree Maps, and more.

Which chart type best fit customer's need and is also suitable for the limited space?

👎 Tree map and Heat map are difficult to display on a limited space. They are also not the best chart type to display the data customers asked for.

👎 Clustered bar chart can fit into the limited space. But will have viewing issue when displaying data across more than 4 years.

👍 Stacked bar chart is the most appropriate chart type.

3. Create detailed designs

Given the limited space available to display the charts, what limitations should be set on the chart display to ensure best usability?

Display limited numbers of category, use “others” to display the rest

How to allow some flexibility for users to create the charts?

For cross tabulation chart, since user need to select two attributes (customer characteristics in MarTech), the order of selection should be flexible
Allows different routes to configure the chart

Segment Insights