Post Modal Redesign
Later
2019
The Later post modal has largely remained unchanged since its original design, with only minor visual tweaks to accommodate new features. This approach has led to accumulating UI inconsistencies and UX debts.
Role
UX Lead
Team
1 UI Designer 1 PM 2 FE Developers
Services
Wireframe Workflow Research Planning
Impact
Auto Publish usage
↑ 4%
Post creation time
↓ 3.5% (9 sec)
Context
Scheduling posts is a critical feature for Later customers, with over 70k weekly posts on Instagram alone in 2019. Content creators, SMBs, and social media managers depend on this functionality to manage their workflows. However, as social media platforms roll out new scheduling capabilities, the current post modal has reached its limits. Adding new features without addressing the existing design will only worsen usability issues, leading to a poor user experience.
This highlights the usability issues and inconsistencies in the UI
Challenges
The primary challenge was managing changes to a core feature for our existing users. From a UX perspective, consistency and familiarity are the keys to a good experience, and a redesign can disrupt that, even if the redesign is objectively better. To address this, I proposed allocating time for user testing early in the process to gather feedback.
Another challenge was creating a scalable design that can accommodate upcoming features. I worked closely with the UI designer to ensure that the redesign used scalable components that could adapt to future needs.
Process
At Later, we use the Shape Up framework, which emphasizes collaboration to understand the problem space and find solutions quickly. As the UX Lead, I pulled everything together into a UX Pitch document, which helped the team narrow down to two directions.
UX Pitch: Low-fi visuals are used to communicate ideas that makes discussion easier
I ranked the post modal features using usage data from Amplitude to see what our customers are doing
I created wireframes for both directions, including images to clearly illustrate the layouts, enabling the UI designer to build the prototype for user testing. Option 2 was preferred due to its large thumbnail and no need for scrolling on desktop, unlike Option 1.
Once the layout was finalized, engineers began scaffolding to prepare a working version for another round of testing. The feedback from this round was favorable, so we proceeded with a percentage rollout, allowing us to fine-tune the product based on real-world use while minimizing user disruption.
Option 1: This uses a table format that is familiar to our users as this is also used in the mobile app
Option 2: This variation tested well in user testing
Features
The publishing method dropdown is now positioned at the top for better visibility. Previously, it was located next to the “save” button, which many users miss.
Captions and additional features are now ordered to follow the users’ workflow from top to bottom. This makes the scheduling process more intuitive.
All buttons have been updated for a consistent look and feel across the interface. Active and hover states have been improved to make interactions more noticeable.
Before & After
Learnings
Early alignment:
Data-driven prioritization:
Value of early user testing:
