MEGAN LIN

HOME

RESUME

ABOUT

CURRENT PROCESS

CURRENT PROCESS

SOLUTION

SOLUTION

RESEARCH

RESEARCH

PAIN POINTS

PAIN POINTS

DESIGN PROBLEM

DESIGN PROBLEM

AviviD.ai | Customer Data Platform

AviviD.ai | Customer Data Platform

Streamlining Audience Segmentation

Streamlining Audience Segmentation

ROLE

ROLE

Product Designer

(solo designer on the team)

Product Designer

(solo designer on the team)

TEAM

TEAM

1 Marketing Specialist

3 Data & Frontend Engineers

1 Marketing Specialist

3 Data & Frontend Engineers

TIMELINE

TIMELINE

December 2024 - May 2025

December 2024 - May 2025

PROJECT STATUS

PROJECT STATUS

Shipped in 2025

Shipped in 2025

CURRENT PROCESS

CURRENT PROCESS

Current customer data platform only supports building audience segments from a brand's own first-party data, which limits customer acquisition.

Current customer data platform only supports building audience segments from a brand's own first-party data, which limits customer acquisition.

Marketing teams at companies like FamilyMart use our CDP to group their existing customers based on contact details and purchase history. This helps bring customers back, but first-party data rarely includes demographics or interests, so marketers know who their customers are but not much about them. Segments are also limited to people already in the brand's database, which makes it hard to reach new customers.

Marketing teams at companies like FamilyMart use our CDP to group their existing customers based on contact details and purchase history. This helps bring customers back, but first-party data rarely includes demographics or interests, so marketers know who their customers are but not much about them. Segments are also limited to people already in the brand's database, which makes it hard to reach new customers.

SOLUTION

SOLUTION

A third-party data marketplace where marketers can purchase segments or attributes to reach new customers or enrich their own first-party data.

A third-party data marketplace where marketers can purchase segments or attributes to reach new customers or enrich their own first-party data.

Select Segment Output

Select Segment Output

Choose between buying 3rd-party segments or enrich 1st-party data with new attributes.

Choose between buying 3rd-party segments or enrich 1st-party data with new attributes.

Match customer IDs

Match customer IDs

Match first-party data to the 3rd-party data to enrich customer IDs with new attributes

Match first-party data to the 3rd-party data to enrich customer IDs with new attributes

Create segments

Create segments

Build audience segments from 3rd-party data

Build audience segments from 3rd-party data

Add audience lookalike

Add audience lookalike

Expand audience by identifying new audience with similar traits and affinities.

Expand audience by identifying new audience with similar traits and affinities.

RESEARCH

RESEARCH

Our users need to attach

Our users need to attach

With PM approval, I interviewed five in-house advertisers: three who use first-party data and two who buy third-party data. I mapped their journeys to surface frustrations, then distilled them into three core insights.

With PM approval, I interviewed five in-house advertisers: three who use first-party data and two who buy third-party data. I mapped their journeys to surface frustrations, then distilled them into three core insights.

PAIN POINTS

PAIN POINTS

No self-serve way to buy audiences

No self-serve way to buy audiences

To purchase third-party segments, marketers had to contact Customer Success and wait to be referred to a partner data provider. This slowed down campaigns and pulled them out of the platform.

To purchase third-party segments, marketers had to contact Customer Success and wait to be referred to a partner data provider. This slowed down campaigns and pulled them out of the platform.

OPPORTUNITIES

OPPORTUNITIES

Bring audience buying into the platform.

Bring audience buying into the platform.

We can build a self-serve Data Marketplace inside Likr CDP, so marketers can browse, purchase and activate third-party audiences without going through Customer Success or outside partners.

We can build a self-serve Data Marketplace inside Likr CDP, so marketers can browse, purchase and activate third-party audiences without going through Customer Success or outside partners.

Design Question

Design Question

How might we help marketers reach new customers who match their best existing one?

How might we help marketers reach new customers who match their best existing one?

Synthesis

Synthesis

Flow Chart with Potential Features

Flow Chart with Potential Features

In order to simulate how users would utilize our platform during the marketing process, I designed a scenario and created a sample user flow that includes proposed feature.

In order to simulate how users would utilize our platform during the marketing process, I designed a scenario and created a sample user flow that includes proposed feature.

1

1

Analyze Trends

Analyze Trends

2

2

Create Segments

Create Segments

3

3

Analyze Segments

Analyze Segments

Sample

User Flow

Sample

User Flow

User Flow

Description

User Flow

Description

Proposed

Feature

Proposed

Feature

Product Filter

Product Filter

Date Range Selector

Date Range Selector

Create charts:

Create charts:

Bar Chart : 40% of yogurt buyers also buy oat milk

Bar Chart : 40% of yogurt buyers also buy oat milk

Demographic Breakdown: Buyers are mostly females aged 25–34 and located in urban areas

Demographic Breakdown: Buyers are mostly females aged 25–34 and located in urban areas

Chart Type Selector

Chart Type Selector

Chart Dimension, Metric, Breakdown Selector

Chart Dimension, Metric, Breakdown Selector

Chart Customization (chart size, placement)

Chart Customization (chart size, placement)

Chart Template

Chart Template

Analytics Insights *allow users to add their own conclusions as notes on the analytics dashboard

Analytics Insights *allow users to add their own conclusions as notes on the analytics dashboard

Database Selection (Sort by Industry)

Database Selection (Sort by Industry)

Segment Accumulation Mode

Segment Accumulation Mode

Hashed Data Mapping

Hashed Data Mapping

Download Audience IDs and Labels *only available for segments that have performed ID mapping

Download Audience IDs and Labels *only available for segments that have performed ID mapping

Traits and Affinities Filter

Traits and Affinities Filter

Audience Lookalike

Audience Lookalike

Segment Preview

Segment Preview

Real-time Analytics

Real-time Analytics

Segment Export to 3rd-party platforms

Segment Export to 3rd-party platforms

Segment Export to AviviD.ai Customer Studio

Segment Export to AviviD.ai Customer Studio

Proposed Feature Color:

Proposed Feature Color:

Delighter Feature

Delighter Feature

Satisfier Feature

Satisfier Feature

Basic Feature

Basic Feature

Proposed Feature Icon:

Proposed Feature Icon:

New feature

New feature

Improved existing feature

Improved existing feature

Pending feature

Pending feature

*Requires further discussion with the dev. team

*Requires further discussion with the dev. team

Select Products

Select Products

Create Charts

Create Charts

Draw Conclusion

Draw Conclusion

Choose Database

Choose Database

Perform ID mapping?

Perform ID mapping?

Build Segment

Build Segment

Check Segment’s

Real-time Analytics

Check Segment’s

Real-time Analytics

FLOW CHART SUMMARY

FLOW CHART SUMMARY

Matching data, expanding audience segments, generating analytics preview could be included in the new create segments flow

Matching data, expanding audience segments, generating analytics preview could be included in the new create segments flow

I flagged features that needed further discussion with the dev team due to potential performance challenges. Tasks like mapping hashed IDs across databases, expanding audience segments, and generating real-time analytics require heavy processing and could increase user wait times.

I flagged features that needed further discussion with the dev team due to potential performance challenges. Tasks like mapping hashed IDs across databases, expanding audience segments, and generating real-time analytics require heavy processing and could increase user wait times.

Customer Studio Analytics

Customer Studio Analytics

Customer Analytics Dashboard

Customer Analytics Dashboard

RFM Dashboard

RFM Dashboard

Customer Studio Campaign

Customer Studio Campaign

Campaign Builder

Campaign Builder

Campaign Table

Campaign Table

Customer Studio Segmemt

Customer Studio Segmemt

Segment Builder (1st-Party)

Segment Builder (1st-Party)

Cross-Platform Syncs Settings

Cross-Platform Syncs Settings

Audience Data Upload

Audience Data Upload

Data Marketplace

Data Marketplace

Segment Builder (3rd-party)

Segment Builder (3rd-party)

Database Selection

Database Selection

Audience ID mapping

Audience ID mapping

Segment Accumulation Mode

Segment Accumulation Mode

Traits and Affinities Filter

Traits and Affinities Filter

Audience Lookalike

Audience Lookalike

Cost Estimation

Cost Estimation

Segment Table

Segment Table

Real-time Analytics

Real-time Analytics

Download

Download

Export to 3rd-party platform

Export to 3rd-party platform

Export to AviviD.ai customer studio

Export to AviviD.ai customer studio

Payment Settings

Payment Settings

Industry Analytics

Industry Analytics

Industry Analytics Dashboard

Industry Analytics Dashboard

Metrics

Metrics

Dimensions

Dimensions

Charts

Charts

Analytics Table

Analytics Table

*Footnote 1

*Footnote 1

*Footnote 2

*Footnote 2

*Footnote 3

*Footnote 3

Current Feature

Current Feature

Proposed Feature

Proposed Feature

Footnote 1:

Footnote 1:

When the user is selecting the audience list for ID mapping, the system will display the list uploaded in the “Audience Data Upload” feature. The list needs to be uploaded prior to it being utilize in the segment building due to the time needed for mapping.

When the user is selecting the audience list for ID mapping, the system will display the list uploaded in the “Audience Data Upload” feature. The list needs to be uploaded prior to it being utilize in the segment building due to the time needed for mapping.

Footnote 2:

Footnote 2:

The user needs to have already synced to the desired 3rd-party platform in the “Cross-Platform Syncs Settings” when choosing to export segments to a 3rd-party platform.

The user needs to have already synced to the desired 3rd-party platform in the “Cross-Platform Syncs Settings” when choosing to export segments to a 3rd-party platform.

Footnote 3:

Footnote 3:

After exporting segments to AviviD.ai customer studio, the segment will be displayed for user to choose from in the campaign builder.

After exporting segments to AviviD.ai customer studio, the segment will be displayed for user to choose from in the campaign builder.

SITE MAP SUMMARY

SITE MAP SUMMARY

Audience matching would be tied to the existing audience upload feature

Audience matching would be tied to the existing audience upload feature

I built a site map to track which new feature would require integration with the existing features.

I built a site map to track which new feature would require integration with the existing features.

Design - “Data Marketplace” Feature

Design - “Data Marketplace” Feature

I. First Iteration - User Flow

I. First Iteration - User Flow

Select Data Source

Select Data Source

Perform ID mapping?

Perform ID mapping?

Yes

Yes

No

No

Upload Audience

List

Upload Audience

List

Do not Audience

List

Do not Audience

List

Set up Segment

Set up Segment

Filter

Filter

No Filter

No Filter

Filter Labels using

Boolean Operators

Filter Labels using

Boolean Operators

Include All Audience

(Skip filtering step)

Include All Audience

(Skip filtering step)

Confirm Segment Size /

Conduct Audience lookalike

Confirm Segment Size /

Conduct Audience lookalike

II. First Iteration - Design

II. First Iteration - Design

Select Data Source

Select Data Source

Set up Segment

Set up Segment

Confirm Audience Size and Cost

Confirm Audience Size and Cost

Segment Audience

Segment Audience

Segment Table

Segment Table

Discussion Snippets - Internal User Testing

Discussion Snippets - Internal User Testing

I tested the design with our in-house marketers and iterated based on their feedback.

I tested the design with our in-house marketers and iterated based on their feedback.

I. Second Iteration - User Flow

I. Second Iteration - User Flow

Original Flow

Original Flow

Select Data Source

Select Data Source

Perform ID mapping?

Perform ID mapping?

Yes

Yes

No

No

Upload Audience

List

Upload Audience

List

Do not Audience

List

Do not Audience

List

Set up Segment

Set up Segment

Filter

Filter

No Filter

No Filter

Filter Labels using

Boolean Operators

Filter Labels using

Boolean Operators

Include All Audience

(Skip filtering step)

Include All Audience

(Skip filtering step)

Confirm Segment Size /

Conduct Audience lookalike

Confirm Segment Size /

Conduct Audience lookalike

Major Bottleneck

Major Bottleneck

Improved Flow

Improved Flow

Select

Data Source

Select

Data Source

Select Goal

Select Goal

Added this step to clarify the workflow’s purpose and route users into the correct path.

Added this step to clarify the workflow’s purpose and route users into the correct path.

Added visualization to explain complex steps (i.e. accumulation mode & audience lookalike).

Added visualization to explain complex steps (i.e. accumulation mode & audience lookalike).

Add Labels to

My Audience

Add Labels to

My Audience

Select Audience List
for ID Mapping

Select Audience List
for ID Mapping

Set up Segment

(Skip filtering step)

Set up Segment

(Skip filtering step)

Build 3rd-party

Segment

Build 3rd-party

Segment

Set up Segment

Set up Segment

Filter Labels using

Boolean Operators

Filter Labels using

Boolean Operators

Confirm All of Segment Info /

Conduct Audience lookalike

Confirm All of Segment Info /

Conduct Audience lookalike

Splits the experience into two paths to prevent confusion.

Splits the experience into two paths to prevent confusion.