MEGA · S4 object storage
How S4 object storage went from prototype to 150TB in one week
I designed S4, MEGA’s S3-compatible object storage experience, making it easier for developers and businesses to manage buckets, permissions, and large-scale data. Through user research, end-to-end design, and close collaboration with product and engineering, I helped shape a simpler, more intuitive alternative to Amazon Web Service (AWS).
Object storage
User research
B2B
ROLE
End-to-end UX & UI
TEAM
TIMELINE
Beta launch to public release
00 SUMMARY
A new S3-compatible object storage product, built on thorough developer testing, and rated 8.2/10 by customers post release.
The Problem
MEGA is best known for consumer (B2C) cloud storage. The new MEGA S4, object storage product needed to earn the trust of developers, sysadmins and business owners (B2B) managing serious workloads, many with large data requirements (25TB+) and years of experience using AWS. The challenge was to make object storage feel simple and more affordable than AWS, without sacrificing the controls and flexibility S3 users expect, within MEGA's current cloud storage product which is built around files rather than buckets.
My role
I designed the S4 web client end to end, covering bucket creation, access keys, user and group management, onboarding, the usage dashboard, account settings, and move and copy dialogs. Throughout, I focused on making S3 workflows intuitive while preserving the security controls users required. The design and decisions were informed by pre-launch user testing I conducted in 2022 (refer to the Key insights section below).
The outcome
150TB uploaded through S4 in the first week across all Pro plans totalling 4% of MEGA’s total upload volume
In a December 2025 survey of approximately 130 S4 customers, satisfaction averaged 8.2/10, with 54% of users classified as promoters. Customers valued MEGA’s trusted brand, and commented MEGA's S4 product had a cleaner interface than AWS, more competitive pricing, and S3 compatibility. The feedback also revealed clear opportunities for improvement: upload speeds, support for large buckets, bucket policies and documentation which formed the second phase's requirements.
01 RESEARCH
Eight research sessions revealed four key use cases, all frustrated by hidden costs and limited documentation
Originally the S4 object storage project was being run by developers, and I was integrated into the project as a junior designer and was the sole designer on the project. Upon seeing the initial product, I recommended running a quick usability test before any further development work was done.
I conducted eight (five existing users and three new to platform) 45-minute usability sessions in March 2022 focusing on understanding users and their key requirements. Each user was a verified developer, sysadmins, or business owner, 60% already used an S3 provider, and most managed 25TB of data or more on the daily. The interviews involved the participant testing a clickable prototype of the product on web client.
Three key research questions
1. Discoverability
Could users navigate from 'Cloud Drive' to 'Object Storage' and create a bucket?
2. Core workflows
Could they understand and complete the flows for managing access keys, users and groups?
3. Language and terminology
Did the content, labels and S3 terminology make sense to experienced users?
Four use cases emerged
Beyond usability, the interviews revealed four distinct use cases, each with different priorities, workflows and barriers to adoption.
Backup and recovery
REQUIREMENTS
Affordable, reliable storage for TBs of server backups, managed through command-line tools.
Pain points
Unexpected retrieval costs made budgeting difficult.
Managing terabytes of data was cumbersome with traditional file or block storage.
Web hosting
REQUIREMENTS
Fast asset delivery, comprehensive API support and reliable vendor support, including integrations with NAS devices.
Pain points
Hidden fees made storage costs difficult to predict.
SDK support was limited, with C++ coverage available, but broader language support lacking.
Machine learning
REQUIREMENTS
Versioned datasets and models, often hundreds of GBs in size. Bandwidth and latency came first, followed by privacy and cost.
Pain points
Documentation focused on only the three leading providers, making alternative solutions harder to evaluate.
The MEGA command-line tool lacked two-factor authentication (2FA).
Outsourcing
REQUIREMENTS
Client-side encryption, straightforward deployment and high bandwidth to deliver S3-compatible storage to institutional clients.
Pain points
Migrating large volumes of client data was a significant undertaking.
A lack of sales and presentation materials made it harder to explain MEGA's value proposition to clients.
A recurring theme: trust
Across these use cases, users wanted more than affordable storage. The key requirements across all four types of users were:
Dependable security
Clear documentation and governance
Predictable costs
Confidence that MEGA could handle their workloads.
This feedback became integral to the next phase of the MEGA S4 product build, and meant developers were building what users actually wanted.
"I can move all my data from places I don’t trust to places I trust a little bit more."
— Research participant, machine learning
Everyone completed every task, but unfamiliar terminology caused friction
All eight participants successfully completed all four prototype tasks, describing the interface as clean, logical and far less overwhelming than AWS.
The biggest usability issues centered around understanding terminology, finding the right entry points and knowing how permissions worked.
What I tested
Create a bucket
All eight completed used the side menu first, rather than the shortcut within the file list
Create a key for a sub-user
Everyone completed the task easily, but “root user” was interpreted differently.
Create a group with a policy
Participants completed the task easily, but wanted supporting links to relevant documentation.
Create a user and attach policies
All participants completed this task, but the relationship between user and group permissions lacked clarity.
Key Findings
Containers in the Cloud Drive list
Finding
Five out of eight participants didn't understand what a “container” was.
Six of eight preferred to keep buckets separate from 'Cloud Drive', with one concerned the buckets would count towards their storage quota.
Action
Keep 'Object Storage' in the side menu
Explain 'containers' through onboarding and documentation.
Clarify bucket and object access
Finding
Users weren't sure how changing bucket-level permissions would affect individual objects.
Action
Clarify the difference between 'bucket' and 'object' level permissions
Show users which files will be affected before changing permission settings.
Improve documentation
Finding
Participants preferred learning through documentation, and often there were no guides provided by other competitors.
Action
Provide create clear, user-friendly documentation.
Expand the help centre with direct access to these user guides for S3 workflows and third-party integrations.
03 design & Delivery
A familiar interface
The final interface lets people create buckets, control access to their data, and share objects by URL, all within the MEGA web client.
After refining the design based on user testing, I worked closely with MEGA’s product and engineering teams to plan implementation and translate the designs into the web app. We worked through edge cases, aligned interactions with user expectations, while maintaining consistency across MEGA’s product suite.
I designed the interface to feel familiar to existing MEGA users, while making object storage easier to understand and manage. This included a clear "Create bucket" action, quick access to bucket settings, and clearer permission controls to help users understand how access is managed at both the bucket and object levels.
Customer feedback reinforced these design decisions, describing S4 as “simple and intuitive” and “easy to use compared to AWS”. We also established ongoing feedback loops, ultimately helping the team monitor adoption, and identify opportunities for future improvements.
Video showing how to create a bucket and the permission flow
Early adoption focused on setup, Pro-plan entry and the dashboard
Guided setup
Most users understood S3 terminology, while others required more guidance to get started. As illustrated by user-research, documentation and guides were key, thereby, to aid early adoption, I designed a step-by-step setup flow covering endpoints, access key creation, and documentation.
A sidebar tracker guides users through three steps: Endpoints, Creating a key, and Creating buckets, as well as having links to the help centre and relevant documentation throughout the process.
Video illustrating the onboarding flow
Onboarding journey flow
Introducing S4 to Pro users
The empty state explains the value of S4 before users enable it.
Users had raised concerns about hidden fees from other providers. I addressed this by highlighting S3 compatibility, 5 times free egress (downloads), and object storage was included in the Pro plan at no extra cost, with an Enable button and clear terms.
The page also provided a direct link to specifications, help articles, and setup guides for third-party tools, including Rclone, Cyberduck and S3 Browser, allowing users to check compatibility and understand how to connect their existing tools before getting started.
Images illustrating the empty state for object storage
Flowchart illustrating enabling/disabling S4
Usage dashboard
I extended MEGA's account dashboard to include S4 alongside Cloud Drive, giving users clear transparency and monitoring of their storage and transfer usage.
The dashboard monitors important metrics such as quotas, base and additional allowances, estimated next-bill costs and daily usage trends. Users can also compare their previous bill with a forecast of the next one.
Image illustrating the S4 dashboard
Video showing the S4 storage and transfer analytics
Supporting flows
I also designed the supporting interactions that help users manage S4 with confidence.
Main account settings
A central place to enable or disable object storage, copy regional endpoints, manage access keys and control thumbnail previews.
Exit survey
Captures why users opt out of S4, helping the team understand friction and identify opportunities enabling increased buy-in.
Move and copy dialogs
Clarifies how data moves between 'Cloud Drive' and 'buckets', with a privacy notice explaining how policies affect access.
04 OUTCOME
S4 became MEGA's fastest growing B2B offering
Within the first week of launching S4 on all Pro plans and improving its discovery within the web client, users uploaded 150TB representing 4% of MEGA’s total upload volume. More than half came from regular Pro users, including 50TB from Pro 3 users alone.
150 TB
uploaded through S4 in the first seven days
4%
of MEGA’s total upload volume in that week
50TB
uploaded by Pro 3 users alone
What customers said
In December 2025, I ran a post-launch survey with approximately 130 S4 users across all Pro plans. This showed a high satisfaction rate for the product, and showed us that backup and recovery were the most common use cases.
8.2/10
average satisfaction across about 130 customers.
42
net promoter score: 54% promoters, 13% detractors.
65%
What users valued
Users highlighted trust in MEGA, value for money, S3 compatibility, and reliable backups as key features which they liked. Many also found S4 simpler to use than AWS, reinforcing the interface's simplicity as a key differentiator.
What needed improvement
Feedback highlighted upload speeds and interface performance when managing large numbers of buckets and objects as key areas for improvement, with the latter receiving the lowest rating of 3.6/5.
"Easy to use compared to AWS."
— S4 customer
"Love having all my cloud tools in MEGA…great to have S4 as part of the MEGA family"
— S4 customer
The survey helped turn customer feedback into clear priorities for future development, and stage two of the project focused on:
Improving upload performance
Making large-scale bucket management more reliable
Simplifying permission controls
Expanding technical documentation.
The team had opportunities to explore these points further with 65% of respondents open to follow-up interviews. This was a great outcome and illustrated a change in project management at MEGA, where previously user research was not integrated into product development.
05 Learnings
What this taught me
What I'd do differently
Document the product to enable collaboration.
S3 and S4 were highly technical, and I was the only designer familiar with the product and its technical dependencies. I would create internal documentation earlier to help other designers understand the product, contribute to the work, and provide feedback throughout the design process allowing for greater collaboration within the project team.
Plan for a longer development cycle.
Building the backend, running beta testing, and launching took around three years, noting, a long beta testing phase was completed due to the complexity of the product. During that time, MEGA introduced a new brand and design system, but S4 couldn't be updated all at once, leaving parts of the interface inconsistent with other MEGA products. For a future product, I would build with shared design system components and tokens from the start, and schedule regular design updates throughout development.
What I'd do AGAIN
Research users before designing.
When I started this project, we had no clear personas or understanding of S4’s target audience. The usability research I conducted uncovered integral users’ needs which were missing from our product prior, workflows and pain points, helping the team define key use cases and identify potential market opportunities.
Design around existing workflows
Research showed that users relied on third-party tools to manage their data. Featuring Rclone, Cyberduck and S3 Browser on the first screen helped users understand how S4 fit into their existing ecosystems. It reinforced that adoption depends not just on the product itself, but on how easily it integrates with the tools people already use.
NEXT CASE STUDY
MEGA
Revamping navigation for 300M+ users
Redesigned MEGA's web navigation to streamline menus reducing cognitive overload, improving discoverability and scalability across MEGA’s growing product ecosystem, without disrupting the workflows of millions of existing users.