← Back
Numerade app icon

Numerade Navigation Redesign

iOS Android Navigation Design Systems User Research A/B Testing

A research-led navigation redesign for Numerade's iOS and Android app — an AI-powered learning platform for high school and college STEM students. From heuristic evaluation through A/B testing to shipped product, the work surfaced core actions in a new top navigation bar and unlocked 175% MAU growth, 3× retention, and 70% revenue surge.

+175% MAU · 60k → 165k in 6 mo
7-day retention
+70% Revenue surge
Credits

Team and role.

Senior Product Designer, Mobile
Heuristic evaluation across iOS and Android, user research synthesis, three competing navigation concepts, A/B test design and analysis with 250+ participants, the shipped top-nav redesign, the six paired problem→solution refactors across the app surface, mobile design system foundations, brand-personality treatment of the home screen.
Engineering
Numerade main page preview — the redesigned home with the new top navigation surfacing core actions
Role Senior Product Designer, Mobile
Platforms iOS · Android
Year 2023 – 2024
The product

An AI-powered learning platform for STEM students.

Numerade helps high school and college students master STEM subjects through an AI-powered learning experience — short video walkthroughs, step-by-step practice, and a growing library of textbook solutions. Used across iOS and Android by hundreds of thousands of learners.

The challenge

An app struggling with engagement and conversion.

Only 24% of users returned after their first session. Existing flows had layered hierarchies, persistent tab bars cluttering the screen, and core features buried where new users couldn't find them. The app had the content; the navigation hid it.

“Make key features easily accessible to improve engagement.”

Process · Research

Heuristics first, then three concepts.

I assessed every critical flow against usability, design consistency, and performance heuristics, partnering with the Director of Product, Lead Developer, engineers, PMs, and researchers. From the audit I shaped three navigation concepts — each a different bet about how users should reach the product's core actions — to put in front of users instead of debating them in a meeting.

Three navigation concepts: Sidebar (Drawer), Dashboard Style, and Step-Driven
Process · Validation

A/B tested before it shipped.

30% Faster task completion
85% Task success · up from 68%
−40% Navigation-related errors
90% Found new navigation intuitive

Participants located features faster — especially the AI Assistant and My Library — and described the new structure as simpler and more purposeful. The test settled the direction with data, not opinion.

The solution

Showcasing what the app could do.

Aligning information architecture (IA) with the user journey. I designed information systems that are user-centered, scalable, and adaptable — built to evolve alongside the people using them.

Numerade top navigation bar — feature previews and platform highlights surfaced directly on the home screen

The top navigation bar carries feature previews and platform highlights into the home view — discovery and depth on the same surface.

Navigation that puts core actions in the home view.

The redesign integrates core actions directly into the home screen via a top navigation bar — replacing a layered drawer with a surface-first model. Users see what the app can do the moment they open it; nothing critical hides behind a menu tap.

Before — the legacy navigation buried core actions behind a drawer

Before — core actions buried behind a drawer menu.

After — the redesigned home with a top navigation bar surfacing core actions

After — the new top nav surfaces the core actions on the home screen.

Solution · Incremental redesign

Six problems, six paired solutions.

Alongside the navigation change, I worked through the rest of the app surface-by-surface, pairing each legacy problem with an explicit design intent — and validating each with engineering, product, and research before shipping. Six surfaces, six before/after pairs.

01 · Problem → Solution

Complex hierarchies → Simplify and empower.

Flattened deep parent-child structures so users could reach what they needed without stepping through layers of context.

Before — complex parent-child hierarchies forcing users to step through layers of context

Before — complex hierarchies.

After — flattened structure that lets users reach their target without intermediate steps

After — simplify and empower.

02 · Problem → Solution

Lack of clear return paths → Instill confidence.

Added consistent back affordances and breadcrumb logic so users always knew where a tap would take them — and how to get back.

Before — missing return paths leaving users uncertain how to get back to context

Before — unclear return paths.

After — consistent back affordances and breadcrumb logic across every surface

After — instill confidence.

03 · Problem → Solution

Overuse of persistent tab bar → Reduce clutter.

Audited the tab bar against actual usage data and removed low-value tabs, freeing screen real estate for the content users came for.

Before — overloaded tab bar with low-value items competing for the user's attention

Before — overused tab bar.

After — streamlined tab bar carrying only the items users actually opened

After — reduce clutter.

04 · Problem → Solution

Insufficient use of modal views → Focus user attention.

Reserved modal views for moments that deserved them — quick tasks, confirmations, focused inputs — so the underlying context stayed intact.

Before — quick tasks pushed users into full-screen contexts they had to navigate back from

Before — modal views underused.

After — focused modal views that handle quick tasks without losing the underlying context

After — focus user attention.

05 · Problem → Solution

Poorly placed features → Improve discoverability.

Moved high-intent features (Ask Educators, Save, My Questions) up the hierarchy where users could see them, not down where they had to dig.

Before — high-intent features hidden several taps deep in the navigation

Before — features buried.

After — high-intent features surfaced on the main views where users can find them

After — improve discoverability.

06 · Problem → Solution

Inconsistent navigation → Build a unified experience.

Standardized navigation patterns across surfaces — same logic, same affordances, same expectations across iOS and Android.

Before — navigation patterns varying across surfaces and between iOS and Android

Before — inconsistent navigation.

After — unified navigation patterns shared across iOS and Android with consistent affordances

After — build a unified experience.

Design system

A shared reference for design and engineering.

The redesign needed a foundation that could outlast it. I developed a design system as a shared reference point for designers, engineers, and stakeholders — improving team productivity and pulling consistency across the app from "intent" to "default." It became the quiet engine behind the user-facing wins below.

Outcomes

What shipped, what moved.

+175%
MAU growth across iOS and Android — 60k to 165k in six months after the redesign shipped. The single number that earned the whole rebuild its budget, and the one a recruiter screenshot of this block alone would communicate.
7-day retention · 1.4% → 4.5%
+70% Revenue surge
3.1 → 4.0 Android App Store rating
3.2 → 4.4 iOS App Store rating
Stack

Tools and methods.

iOS Android Heuristic evaluation A/B testing Design systems User research Cross-functional WCAG AA

“One structural change — surfacing core actions in the home view — compounded across retention, ratings, and revenue. The biggest lesson wasn't the design itself; it was the discipline of testing the choice before believing it.”

Visit numerade.com