background

Creating an educational popover took two months

EHQ is Pinterest's self-serve platform for building in-app experiences and tracking growth metrics. Teams used the platform to build educational popovers to introduce new features, but the creation process was complex and time-consuming.‍

The hypothesis

We thought anchor selection was the biggest source of friction

Early assumptions identified anchor configuration, the step where users select where a popover appears across surfaces (e.g. Homefeed) as the primary point of friction.

The initial goal was to simplify this step and reduce setup time by 85%, but research later revealed gaps in that hypothesis.

research

Four friction points were slowing teams down

I partnered with another Product Designer to conduct contextual inquiries with Product Marketing Managers and Marketing Campaign Specialists. Together, we synthesized the findings into four key pain points that informed the concepts I explored.

Pain point #1: Configuration was difficult to understand

Technical jargon, an unintuitive layout, and inconsistent terminology made it difficult for users to configure educational popovers.

Pain point #2: Validation was inconsistent

Inconsistent required field indicators across the Create modal and Configuration form caused users to miss steps and resulted in broken live experiences.

Pain point #3: Teams couldn't predict in-app placement

Users lacked confidence in where anchor options would appear in the Pinterest app, leading to guesswork and edits after publishing.

Pain point #4: Creating for multiple platforms required duplicate work

Teams had no streamlined way to submit the same educational popover across iOS, Android, and Web within a single form.

Reframing the problem

The challenge was bigger than anchor selection

💭 How might we: Enable teams to create educational popovers independently and minimize setup errors?

Contextual inquiries showed that anchor selection was only one part of the problem. Teams also struggled with jargon, inconsistency, uncertainty about placement, and rework.

Ideation was collaborative, while I owned the following explorations:

solution highlights

Making educational popover creation easier to complete independently

Keeping creation and configuration in one flow

I moved creation from a modal to a full-screen experience with a 'continue' button, allowing users to move directly into configuration rather than returning to the list view.

Replacing technical language with familiar terminology

The original form was designed around developer needs but used daily by product and marketing teams. I restructured the layout and replaced technical labels with language that aligned to users' workflows.

Making relevant anchors easier to find

I redesigned the anchor dropdown to support keyboard search and exclude inactive options. We also explored automating anchors across operating systems to eliminate manual engineering work.

Giving teams confidence before publishing

I designed a placement preview so teams could see where educational popovers would appear in the Pinterest app. Technical constraints prevented a live preview, so I created a simplified representation that still gave users confidence in their selections.

Modernizing the experience

I used Gestalt components to improve consistency and accessibility while maintaining workflow-specific interactions.

outcomes

The redesigned creation workflow is expected to reduce set up time to less than two weeks. Stakeholders found the simplified field labels, searchable dropdowns, and new layout more intuitive, empowering teams to create educational popovers independently and more quickly.

While my contract concluded before release, I partnered closely with another designer to co-lead discovery and I owned key explorations that informed the final product direction.