Login Redesign

The project had stalled for two years. A mockup I made revived it.

As the platform expanded its cloud offerings, the login experience became the primary entry point for customers accessing multiple products and tenants. Functionally, it worked — but visually and behaviourally, it lacked consistency, clear messaging, and polish.

The project had originally stalled in 2023 due to undefined requirements. Two years later, a simple mockup I created sparked renewed interest. I was tasked with a refresh — but I quickly identified that a visual update wouldn't suffice. The flow had logical gaps, ambiguous feedback, and inconsistent branding. I took ownership of rebuilding the login experience from the ground up: intuitive for users, rock-solid for engineers, and aligned with the Shield Design System.

Five pain points that made this more than a refresh.

User feedback, internal reviews, and my own audit surfaced critical problems across the authentication journey:

Five targeted refinements that rebuilt trust in the system.

I approached this not as a single "redesign" but as a series of targeted refinements addressing each problem at the root cause.

Writing SDS-5676 and driving the texture-only pivot.

The second, more strategic layer of this project involved the login page's visual background — and a trust problem that went beyond copy and layout. User feedback had surfaced concerns about the existing side imagery: it felt dated, didn't align with the brand's evolution, and for security-conscious admins, the sudden URL change combined with an unfamiliar-looking page was raising phishing red flags. An engineering ticket explicitly called out that users were mistaking the new login page for a phishing attempt.

I authored SDS-5676, the Shield Design System story for updating the login screen background. I framed the requirements around six specific criteria that would make the login page durable and trustworthy:

During the cross-functional design review, I participated alongside the design lead, a product designer, and an engineer working through two layout options: texture-only versus background with a side image. I flagged that the side image approach carried compounding risk — imagery ages, requires repeated update cycles, and introduces visual noise that competes with the one thing users need to focus on. The decision the team agreed on was to remove the side image entirely and proceed with texture-only. This wasn't a rejection of the image concept; it was a deliberate deferral pending Marketing's brand direction confirmation.

I captured the full decision rationale in the Design Decisions Log — a structured documentation system I created to make design rationale discoverable and auditable for cross-functional decisions.

Two lessons that changed how I approach flows.

I initially treated the login screen like a standard form — using inline errors. I quickly learned that authentication errors come from the backend, not the UI, which dictated a shift to banner-level alerts. It taught me that design must reflect system architecture, not just visual conventions.

The project didn't stall because of bad design — it stalled because of undefined behavior. Once I mapped out the logic, edge cases, and error states, the path to production became clear. This reinforced that my role isn't just to make things look good, but to untangle complexity so the team can ship.

Why the login page matters at scale.

• Reduced trust erosion at the authentication surface — Every activated customer's first platform interaction is this page. Moving from stale imagery to stable, brand-consistent texture directly reduces the phishing-perception risk that was documented in support tickets.
• Reduced support case volume — Clearer status messages, explicit button labels, and better error feedback addressed recurring "stuck login" confusion patterns from the Platform Support FAQ.
• Activation velocity — With major products targeting cloud-only by Q2 2026 and expanded activation in Q3, every improvement in first-login success rate compounds across thousands of activating users.
• Competitive parity — CyberArk and peer enterprise security companies use clean, texture-based login treatments. The previous approach was visibly behind market standard for an enterprise security product.
• Design system durability — The texture-only approach means no recurring update cycles for imagery. One design, one token set, minimal maintenance — freeing the team for higher-impact work.

Tracked, documented, and handed off.

Not everything shipped in this cycle. The following items were documented and tracked before handoff: