Startup Event Organizer

Designing Scalable Segmentation Tool for High-Volume Audience

From Draft to Done,
Redesigning an AI Editor
Researchers Stay In

Company

Confidential

Timeline

1 week

Role

UX Designer

Scope

Design System
Interaction Design
Stakeholder Management

This is a design brief from a B2B organisation in the events industry. The project goal is to build an internal notification segmentation tool that allows relationship managers, marketing, and product teams to create targeted notifications across multiple channels without requiring technical support.

Business Problem → Design Goal

Too Many Notifications, Not Enough Relevance

Untargeted broadcasts

Notifications were sent to all users with minimal filtering, regardless of relevance to the individual recipient.

Design Goal 01

Flexible Segmentation

Let teams build audience segments based on company type, location, industry, and lifecycle stage

which leads to…

Notification fatigue and low engagement

The lack of targeting led users to tune out or disengage from notifications over time, diminishing their effectiveness as a channel.

Design Goal 02

Intentional Targeting

Reduce noise for end users by making targeting intentional and data-driven

and causes…

Missed timely opportunities

Without precise segmentation, the organization couldn't reliably reach the right users at the right moment, causing relevant messages to get lost in the noise

Design Goal 03

Unified Workflow

Enable notification creation, preview, and scheduling within the same tool

Design Decision

Cross-condition Segmentation Creation

The biggest challenge was designing for the variability of segmentation logic: the combinations of IF, AND, OR, and ELSE users might need. Segmentation originally covered only profile attributes, so I added behavioral data as a second category to capture timeliness and relevance.

Each filter lives as its own label, visible in the same space, so users combine conditions across both categories. This also makes adding new filter more scalable, as it only requires adding a new label.

Design Decision

A Guided Flow for Real Notification Needs

Creating a campaign involves three distinct tasks: selecting recipients, composing the message, and configuring delivery, so I broke it into a 4-step flow instead of one overwhelming screen.

For the message step, I went beyond the brief and added a link field, since notifications usually also meant to drive action, and teams need to direct recipients somewhere specific. This also opens the door to UTM tracking, which marketers rely on to measure click-through performance.

Design Decision

Progressive Scheduling with Built-in Safeguard

Scheduling needed to support two modes: time-based and behavior-triggered. Within time-based scheduling, I restructured one-time and recurring options into collapsible sections so users commit to a mode before seeing its configuration.

I also added a frequency capping rule, beyond the original brief. Since multiple team members can independently create campaigns, the same end user could receive several notifications in a day from unrelated campaigns. The cap acts as a system-level safeguard against that.

Key Insights

Building the Design System from the Bottom Up

Two-layer components, built from the smallest unit. I started from the smallest components rather than full screens, forcing structural decisions early rather than as expensive fixes later. I split them into a universal design system for reusable components and a project-specific layer for this tool's context, keeping the shared library clean while giving room to be specific.

Designing as stress-testing. Even with a well-defined structure upfront, edge cases and structural tensions don't always surface until the design is actually in front of you. The act of designing became its own way of testing whether that structure held up.

Next Case Study

Previous: Aveksana AI

Aveksana

Previous: Aveksana AI

Redesigned AI editor page, onboarded 3 pilot B2B users

Previous: Aveksana AI

IA

Previous: Aveksana AI

Design System

Previous: Aveksana AI

Stakeholder Management

Previous: Aveksana AI

Rocky Paroky

Web Designer & Founder

Book a Call

Almost There!

work@rockystudio.com

Made with Framer, powered by snacks and love.

I approach problems with structures, but solve beyond them.

©2026Felicia Margaretha Karso