Redesigning Megaputer: Turning a 100+ Page Information Maze into a Trustworthy, Navigable Product
Megaputer is a no-code platform for data and text analytics — powerful software buried inside a website that users couldn't navigate, didn't trust, and couldn't make sense of quickly. Over a year, our team redesigned the entire experience from the ground up: restructuring a fragmented information architecture through real card sorting sessions, rebuilding the visual language from scratch, and validating every major decision with the same users throughout. The result was a site that users could actually use — and one that reflected the credibility of the product behind it.
Read the Full Case Study Below
What is this project about?
UX Optimization
Redesigned a B2B data analytics website to fix a broken information architecture, reduce content overload, and make the product actually understandable within seconds.
Research & Testing
Recruited 10-20 real users including actual clients for interviews, card sorting, A/B testing, and before/after usability testing, nothing was assumed.
Design Execution
3-4 rounds of wireframes, 2-3 rounds of high fidelity UI, and a design system built from scratch to bring consistency to a fragmented visual experience.
Client:
Megaputer, a no-code platform for data and text analytics, aimed to redesign and rebrand its website to improve user experience and better reflect its brand.
Project Overview
⏳ Timeline
- 1 Year
- Group Project
👩💼 My Role
- UX Research
- Ideation & Brainstorming
- Hifi Designs & Prototyping
- Usability Testing
🎉 The Result
- 90 seconds → 25 seconds task completion
- 127 pages restructured to 50
- 9 out of 10 users preferred the redesign
📚 My Learnings
- How to untangle a deeply broken information architecture
- Making design decisions based on user data not assumptions
- Iterating through multiple rounds without losing sight of the core problem
What was the main problem?
Nobody Could Find Anything
The site had 127 pages, deep subpage nesting, and duplicated content scattered throughout. Users had to click through multiple levels just to find basic info, most gave up before getting there.
The Site Didn't Feel Trustworthy
Inconsistent UI, outdated imagery, and unclear CTAs made the platform feel unpolished. For a B2B product selling to enterprise clients, that lack of credibility was directly hurting conversions.
Too Much Text, Too Little Clarity
Every page was dense with content but light on structure. Users couldn't skim, couldn't prioritize, and couldn't quickly understand what Megaputer actually offered or whether it was relevant.
Reworked the existing Information Architecture to expose all the hidden issues – the current IA of the website! Look how massive it is !! 😱
Impact of Redesign
Clearer Information
127 Pages → 50 Pages
Streamlined Information Architecture Card sorting sessions with real users, drove every structural decision. The result was a navigation built around how people think, not how the product was built.
Faster Task Completion
90 seconds → 25 seconds
Task Completion Time Measured through before and after usability testing with the same participants using identical tasks. Users completed core workflows nearly twice as fast after the redesign.
Preferred the Redesign
9 out of 10 users
Preferred the Redesign Tested with the same participants before and after — 9 out of 10 said the new design was clearer, faster to navigate, and more trustworthy than the original.
How did we address the problem?
The Solution Break Down
Restructuring the Information Architecture
The existing site had 127 pages, many duplicated or buried under multiple levels of navigation. We ran card sorting sessions in Miro with real users — including actual Megaputer clients — to understand how they mentally grouped content. Based on those findings, we restructured the entire site down to 50 pages, organized around how users think rather than how the product was built.
Building a Design System
Before designing any screens, we established a unified component library, typography, colors, buttons, cards, spacing. This ensured every page felt consistent and made iterating across 3-4 wireframe rounds and 2-3 UI rounds manageable without starting from scratch.
Reducing Content Overload
We introduced collapsible dropdowns across content-heavy pages, letting users reveal only what's relevant to them. This cut visual noise significantly without removing information that some users still needed.
What else did we improve?
❌ Before
The old hero section confused users with unclear imagery and tagline, failing to convey the company's purpose.
✅ After
The new hero section clearly communicates the company's purpose and highlights trusted partnerships, proper imagery boosting credibility.
❌ Before
Core offerings were buried inside a tab, causing user confusion.
✅ After
Key offerings are now highlighted on the homepage for easy access.
❌ Before
Users struggled with too many categories and a lack of clear descriptions, making it difficult to find relevant solutions quickly.
✅ After
Streamlined categories (35 to 21) with tab-based navigation, contextual tags, and a clear tagline for better usability and SEO.
❌ Before
Product pages were walls of text with no hierarchy, users couldn't skim or quickly grasp what each product did.
✅ After
Introduced product cards with concise descriptions and clear CTAs, making it easier for users to skim and grasp key details.
🎉 New
A comparison with competitors that emphasizes the company's value proposition and boosts SEO with targeted keywords on the homepage.
🎉 New
Added breadcrumbs for easier navigation, allowing users to understand their location within the site and improving overall usability.
How did we address the problem?
01
Research & Discovery
We recruited 10-20 participants through LinkedIn and personal networks: including actual Megaputer users, to understand how they experienced the site. The findings were consistent: users couldn't tell what the company did, couldn't find what they were looking for, and didn't trust what they saw. Too many pages, too many clicks, duplicated content, and pictures that felt unpolished.
02
Card Sorting | The Hardest Part
The information architecture was the core problem. We ran card sorting sessions in Miro, asking participants to group and place content where they naturally expected to find it. Instead of assuming how pages should be organized, we let users tell us. This process took days — the existing IA was so fragmented that even sorting it was a challenge. But the output was a structure built on how people actually think, not how the product was built.
03
Design System
Before touching screens, we built a design system from scratch. Consistent components, typography, color, and spacing, so every page felt like it belonged to the same product. This took significant time upfront but made every design decision after it faster and more consistent.
04
Wireframes & Iterations
We went through 3-4 rounds of wireframes before touching visual design, then 2-3 rounds of high fidelity UI. Nothing was finalized without going back to users first.
05
A/B Testing
One specific challenge was information density — the site had too much content and we couldn't agree on the right layout direction. We ran an A/B test on LinkedIn to let real users decide. That result informed the final content structure.
06
Usability Testing — Before & After
We tested with the same users before and after the redesign, giving them identical tasks each time. Before the redesign, users averaged 90 seconds to complete core tasks. After, that dropped to 25 seconds. That's a real, measured improvement — not an estimate.
The Redesigned Screens
Every screen below solves a specific problem we found during research.
The consistency you see across pages isn't accidental — it's the design system doing its job.
Each screen addresses a specific problem we found during research. The homepage communicates what Megaputer does within seconds — something users consistently failed at before. Solution pages are reorganized with tabs, tags, and short taglines so users can filter and scan instead of clicking blindly. Product pages replace text walls with scannable cards and clear CTAs. Resource and blog pages are restructured so content doesn't get buried. The trial and contact pages are simplified to reduce friction at the most important conversion points. The consistency across all of them comes from the design system we built before touching a single screen.
What I took away from this project
Real users change everything
We could have assumed how to reorganize 127 pages. Instead we ran card sorting with actual users and the structure they gave us was completely different from what we would have designed on our own. That gap between designer assumptions and user reality is exactly why research exists.
A/B testing saved us weeks of debate
We were stuck on information layout for too long, too much content, two valid directions, no clear answer. Putting it in front of real people ended the debate in days. When you can't decide, test.
The design system was the best time investment we made
Building it before touching screens felt slow at the time. But going through 3-4 wireframe rounds and 2-3 UI rounds with a shared component library meant we were iterating on decisions, not rebuilding elements. Every hour spent on the system saved three hours later.
