Iterable: Designing a Site Structure That Could Survive a Rebrand

Overview

Role: UX Designer (Design Systems)

Timeline: Jan to Apr 2026

Scope: Owned IA, wireframes, and component architecture solo. A UI designer applied the rebrand's visual design on top of my structure.

TL;DR: The rebrand's visual direction was still moving, so I separated structure from styling. That let IA and components progress without waiting on the brand, and let the UI designer reskin without me rebuilding anything.

Problem

I joined Iterable while a full rebrand was underway, which meant the visual identity was unsettled while I needed to establish the site's structure. A UX audit also surfaced a dead conversion journey: competing calls to action with no clear hierarchy, and social proof buried below the fold, too late to influence a decision.

The decision: structure first, style later

If I had built components tied to the current visual direction, every change in the rebrand would have forced a rebuild. So I made the component library brand-modular: layout and structure logic decoupled from visual styling.

The tradeoff was upfront abstraction. I had to define what a component was structurally before anyone could show me what it looked like. In return, my work was never blocked by the brand, and the UI designer could apply the evolving visual system without waiting on my components to be "finished."

How it came together

I audited the existing site, built a new IA aligned to Iterable's product, solutions, and resources hierarchy, and wrote wireframes for the priority pages. In the component library I refined existing components and created new ones where none fit: heroes, feature grids, trust elements, and resource blocks. The UI designer built the high-fidelity visual design directly on my wireframes.

The dead journey

I diagnosed the conversion problem as structural, not aesthetic: the page gave users no clear primary action and showed proof too late. I took the finding to the director, got buy-in, and led the restructure within the new IA.

Outcome

The journey that had been dead was no longer dead once the new structure shipped.

I don't have conversion numbers for this fix. The contract ended before that data was visible to me, and I'd rather say that than estimate one.

What I'd do differently

The contract was short and the brand was in flux, so I didn't build a contribution or governance model for the component library. With more time, that would be next. It's what turns a well-structured library into one other people can safely extend.