Helping Marketers create campaigns with complex data

Role: UX Designer
Project Length:
2 months
B2B Software new feature design
Beta feature in Early Access


During my internship at Braze, I designed a beta feature to help Marketing teams create more accurate targeted campaigns with complex data. Braze is a Customer Engagement software platform used by businesses for cross-channel marketing (email, SMS, In-App messages). For this project, I was on the Data Ingestion team.

This feature was in Early Access, and had a low adoption rate. Due to confusion with the current solution, customers were creating incorrect marketing segments, targeting the incorrect users in campaigns, and wasting marketing budget. Additionally, Braze Sales teams were facing challenges addressing this limitation in deals versus competitors.

Sample screenshot of the platform from the Braze website. Users can create marketing campaigns with a variety of filters and data gathered from customers.


Initial research was conducted through internal interviews with the Product, Solutions Consulting, and Sales teams combined with insights from competitive research.

Screenshare from an internal user interview with a Product Manager, demonstrating how creating a campaign with complex (nested) data was confusing and unclear.

Competitive Research in addition to existing documentation. I also searched for product demo videos, third party reviews, and support documentation since it was difficult to obtain logins for competitor software without being a paying customer.


It was especially helpful to do early sketching sessions with the product manager, other designers, and engineers. These were also good in understanding how nested data works, which is a main part of the feature I was designing for. During these sketching sessions, I could also validate ideas for technical feasibility and scope out potential limitations with the larger team.


Sketches from collaborative sessions helped me create several low fidelity wireframes to present during our next design review. To help paint the picture, I based these wireframes on realistic customer scenarios.

For example, how would a marketer using Braze target users who've been on a basic subscription plan for more than 2 months with a limited time promo for an upgraded subscription? How would they use nested data attributes to create a accurate campaign? These were some of the scenarios I tried imagining from the customer's perspective.


After several internal design reviews, I created high fidelity mockups and prototypes based on feedback.

There were two main versions that we wanted to test with customers and validate. The first concept was around exploring data attributes by expanding and collapsing menus, and the second concept focused more on a search experience where the user could find exactly what they were looking for.

Since Braze users have varying levels of technical expertise, we were seeking clarity around how users wanted to find the desired data attributes they wanted for their marketing campaigns. Non technical users might want to explore the data structure, while more technical users might be more comfortable searching for these data attributes directly.


While scheduling usability interviews with customers, I also gathered feedback from Product and Engineering to continue iterating on designs before testing. Self-recorded videos were especially helpful in gathering asynchronous feedback.

I also created a user interview script for guiding moderated user testing with customers, and led these interviews while observing how customers interacted with the prototypes.

Customer interview where a user was going through the flow of exploring their data structure.

Final Designs

The majority of feedback was centered around combining the explore and search interfaces into one experience. This was our hypothesis, but it was helpful to get customer validation to back up design decisions. Considering the varying technical skills of users, it made sense to provide a flexible experience of exploring and searching for data attributes.


After our findings were synthesized, a Jira ticket was created for the lead engineer to start developing based on these designs. I also added annotations and guidelines to the designs before hand off, while referencing components in the design system that could be used. This was right around when my internship was ending, so I could not see the implementation of these designs in the Braze platform.

Back to Top