How Splitting Login and Signup Increased Activation by 12%

The Problem We Saw

The neobank client had a unified auth screen: email input, then a toggle between "Don't have an account? Sign up" and "Already have one? Log in." On mobile, this pattern meant users entered their email, tapped the toggle, and form fields reorganized. It felt confusing. Data showed 18% of users abandoned at the auth step.

We hypothesized that splitting into two clear screens—one for signup, one for login—would reduce cognitive load and increase completion.

The Test Setup

We ran a 2-week A/B test with 50/50 traffic split:

Control (Unified): Email input. "New user? Click here" toggle. Form adapts to signup or login mode.

Variant (Split): User lands on a "Sign Up" button and a "Log In" button. Tapping either takes them to a dedicated screen with one clear path.

Both flows converged at the same KYC step. The test ran across iOS and Android simultaneously.

What Changed

Signup Screen (Variant)

• Large primary button: "Create Account"
• Email field with label "Email address"
• Password field with inline strength indicator
• No secondary action; full focus on signup
• Copy: "First time? Let's set up your account" (reassuring tone)

Login Screen (Variant)

• Large primary button: "Sign In"
• Email and password fields, stacked clearly
• "Forgot password?" link at the bottom
• No signup option visible (reduces friction for returning users)
• Copy: "Welcome back. Sign in to your account"

The First Screen (both variants)

This was critical: we added a simple decision screen before directing users to signup or login. Two buttons. That's it. No text, no forms yet. This gave users a moment to orient themselves.

The Results

Activation Rate

• Control (unified): 67% completed account
• Variant (split): 75% completed account
• Lift: +12%

Drop-Off at Auth Step

• Control: 18% dropped at auth
• Variant: 8% dropped at auth
• Reduction: 10 percentage points

Time to Complete Auth

• Control: 2m 14s average
• Variant: 1m 52s average
• 22-second improvement (17% faster)

Login vs. Signup Behavior

Interestingly, 64% of users tapped "Sign Up" and 36% tapped "Log In." This ratio remained consistent across Android and iOS, suggesting genuine split usage.

Why It Worked

Three things:

1. Clarity of Intent — Users knew immediately what they were doing. No parsing a toggle or re-reading copy. Separate screens = separate mental models. One screen, one job.

2. Reduced Cognitive Overhead — Toggling between signup/login modes requires working memory. On mobile, where attention is fragmented, this friction compounds. Split screens eliminate it.

3. Mobile-First Design — Toggles work fine on desktop with a cursor. On mobile, they're smaller, require precision, and add a step. Buttons are bigger, easier to tap, faster to process.

Implementation Notes

We kept these principles:

• One decision screen first: Two buttons, nothing else. Gives users a moment to choose.
• No back button fatigue: We let users swipe back to re-choose (common iOS pattern) instead of a visible back button.
• Password recovery visible: We kept "Forgot password?" on the login screen, not hidden in a menu.
• Consistent branding: Both screens used the same typography, colors, and spacing. Felt cohesive despite being separate.

The Lesson

Unified flows sound elegant. They're often worse. Users come with one goal: sign up or log in. Not both simultaneously. Designing for that single, clear goal beats designing for hypothetical flexibility. The 12% lift justified the extra screen and made the product feel faster, simpler, and more trustworthy to new users.