Reducing Trial-to-Conversion Drop-off

Redesigning a post-trial payment experience to increase deposits for a leading spiritual advisor marketplace.

B2C . CONVERSION

What's this about?

Problem . My role . Solution

Problem

We assumed customer was ready to commit

Despite online chat being the preferred method for new customers to contact an Advisor on Keen, a marketplace owned by Ingenio, a significant trial-to-conversion drop-off occurred across all platforms.
58%
Trial-to-conversion drop-off across web and mobile app platforms
My Role
MY ROLE
Lead
Senior Product Designer
TIMELINE
3 months
Concept to release
PLATFORM
Mobile app
+ Responsive web
COMPANY
Keen
e-comm marketplace
Solution

Reducing cognitive load via visual/UI optimization at the moment the trial period ends.

1
I dug deeper to find the real problem
The design treated users like they'd already decided to pay, when they were still figuring out if they could trust the service.
2
I worked through the problem layer by layer
Instead of jumping to visual fixes, I started with what users needed (to feel confident, not scammed), prioritized which friction points to tackle, then focused my designs to reduce cognitive load.
3
I made targeted changes within real constraints
I couldn't change the timer, backend, or pricing. So I focused on what I could control: keeping users in the chat, reducing copy, showing fewer options, pre-selecting the lowest amount, and making buttons easier to tap.
IMPACT
+9.5%
Mobile app deposits increase
Before

High Friction at the Critical Moment

What users experienced:
  • Users still building trust but treated like committed buyers
  • Full-screen modal took users out of the chat (mobile app)
  • 4 payment options with dense text to process within 60 seconds
  • Touch targets too small for mobile
CONVERSION PATTERN
58% drop-off at trial expiration.
After

Producing more confident decisions

What we changed:
  • Bottom sheet kept users in the chat while choosing
  • Reduced to 3 options with clearer visual hierarchy
  • Pre-selected the lowest amount as default
  • Increased touch targets for easier tapping
  • Dynamic CTA showing exactly what they'd get (e.g. Add 10 more minutes)
KEY OUTCOME
+9.5% mobile deposits

Cognitive load reduction worked where users already had intent—mobile app users were ready to pay, they just needed a clearer path.

How did we get there?

Problem . Research . Framing

THE CONTEXT

Trust Barriers in a High-Stakes Category
Talking to psychics can be fraught with real-life barriers, but paying to talk to one online adds even more barriers to trust.

THE TRUST CHALLENGE

Psychic services face inherent trust barriers that compound at the payment moment.
Category skepticism
Users often come in with baseline doubt about legitimacy
Payment vulnerability
Asking for money amplifies existing trust concerns
Time pressures
60-second window creates urgency without confidence
Business Model
Converting trial users ("5 minutes for $1") into paying customers to justify acquisition costs.
THE PARADOX
The moment users are asked to pay is the exact moment they're still forming trust—creating a fundamental tension in the conversion flow.

THE RESEARCH

When Trial Users Become Paying Customers
The moment a user's trial ends is the most critical point in the conversion funnel.
Why online chat matters
Online chat was the preferred method for new customers to contact an Advisor—making this the primary conversion path.

This was due to:
Lower barrier to entry
Chat feels less committal than a phone call for first-time users
Privacy and control
Users can engage at their own pace without verbal commitment
Mobile-first behavior
Most users discovering Keen are on mobile devices
A closer look at this area revealed:
Patterns (Trends)
Mobile consistently 30% lower conversion than web across all user cohorts; pattern exists regardless of advisor quality or time of day
Mental Model
Original design assumed trial users were already committed buyers who just needed payment options

DATA ANALYSIS

Three Critical Exit Points
Further user flow analysis revealed the specific causes for user abandonment, particularly in mobile experiences.
Exit points
Exit point 1
Context Disruption

Full-screen modal removed users from chat, breaking the connection with their advisor at the moment of decision.
Exit point 2
Cognitive Overload

4 visually similar options + dense copy + disclaimer exceeded working memory capacity under 60-second time pressure.
Exit Point 3
Decision Paralysis
No default selection + unclear value hierarchy caused analysis paralysis, leading to timeout.
THE CASCADE EFFECT
Each issue amplified the others. Context disruption made cognitive overload worse, which made decision paralysis more likely, which led to timeout abandonment.

THE OPPORTUNITY

Framing my approach
Given what was uncovered, I created a strategy map to help guide my approach and to better articulate to stakeholders the direct line from design decisions to business outcomes.
IDENTIFY...
  • Ways to increase credibility during payment decision
  • Ways to improve usability under time pressure (60-sec constraint)
  • Ways to reduce uncertainty about cost commitment
WHICH ENABLES...
  • Increases trial-to-paid conversion
  • Reduces customer acquisition cost (higher trial ROI)
AND DRIVES...
  • More incremental annual revenue from mobile users
  • Improved customer lifetime value through better first-impression conversion
IN SUM...
If we increase users' confidence and reduce cognitive load during payment decisions (desirability), then we increase the rate at which trial users convert to paying customers (customer acquisition), which directly increases revenue and improves competitive positioning (financial).

What was the approach?

Ownership . Prioritizing . Validation

MY APPROACH

End-to-End Design Ownership
I owned the complete design process to encourage users to pay at the moment their trial period ended.
Structural Issues Identified
I facilitated several meetings to move beyond surface symptoms and uncover the root cause.

Which was the payment flow created a cognitive load spike at the critical trust-formation moment.

What led to that was:
⏱️ 60-second decision window
Temporal constraint created artificial urgency
📋 4 payment options with dense text
Choice overload exceeded working memory capacity
📱 Full-screen modal takeover
Context disruption broke advisor connection
⚖️ Legal disclaimer required
Additional cognitive burden
DESIGN CHALLENGE
How might we reduce the cognitive load on the user by streamlining the UI so customers can make their choice with confidence before the time runs out?

DESIGN PROCESS

From Constraints to Confidence
By systematically working through the constraints, and priorities, I delivered a mobile-optimized solution focused on cognitive load reduction.
Constraints
Balancing between 3 forces
Business Constraints
  • Must increase conversion without changing pricing model or payment infrastructure
  • Cannot extend decision time beyond 60 seconds (legal/product requirement)
  • Success metric: Measurable lift in deposit completion rate within 3 months
  • Budget: Limited to design/frontend changes only
Technical Constraints
  • Cannot modify backend payment logic or pricing tier algorithms
  • Must work within existing A/B testing framework
  • Mobile app platforms (iOS/Android) required different implementation than web
  • Timeline: 3 months to ship
User Constraints
  • Users needed clarity under time pressure (60 seconds to decide)
  • High trust barrier: paying real money to psychics is inherently high-friction
  • Mobile users in thumb-zone ergonomics context
  • Preferred chat (80%) but chat also had higher abandonment risk
Strategy
Setting Prioritization By Using Pain Point Matrix
KEY DIRECTION
This framework justified focusing resources on mobile-native patterns (easy + high impact) rather than attempting backend changes (hard + high impact).
Design
Designing To Address The Highest Priority Solutions
Reminder
This was what it was before.
1. Decision paralysis (no default)
→ Solution: Pre-select lowest tier
2. Context switching (taking user out of chat)
→ Solution: Bottom sheet pattern (iterations shown below)
3. Visual clutter (dense copy, similar options)
→ Solution: Streamlined UI, copy reduction
4. Mobile touch targets too small
→ Solution: Increase to 48px minimum (iterations shown below).
Other callouts
Improving the Information Architecture Using UX Laws
Jakob's Heuristic #1 — Visibility of System Status
Timer remained visible = clear feedback on decision window.
Jakob's Heuristic #6 — Recognition Over Recall
CTA copy dynamically showed time added = immediate value recognition. Example:
Summary
Design Decisions
Minimal visual noise
Removed elements that didn't serve decision-making
Typography hierarchy
Clear distinction between price, time, and action labels
Brand color on CTA only
Drew attention to primary action without overwhelming
Subtle elevation
Bottom sheet shadow created clear hierarchy over chat

VALIDATION

Reducing Cognitive Load
The hypothesis was that users weren't opposed to paying—they were overwhelmed by the ask at a vulnerable moment.
Validation method: A/B testing across platforms
To validate the cognitive load hypothesis, I worked with our lead data analyst to design a controlled A/B test comparing the original experience against my optimized designs.
Test structure
Control (A): Original payment flow
Variant (B): Optimized for cognitive load
Platforms: Mobile app + Responsive Web
Key metric
Deposit conversion rate
Percentage of trial users who successfully add funds
A/B Testing Framework Limits
Working within the existing testing framework meant several variables were locked:
Cannot change
  • Time users had to make a choice (1 minute)
  • Number of options presented
  • Backend logic for price bucket options or min/max custom amounts
My strategic bet
Focused on mobile-specific UI/UX optimizations requiring no backend changes while maximizing behavioral impact—specifically reducing cognitive load via mobile-native design patterns and streamlined information architecture.
Then a change of scope happened
This constraint analysis justified our VP of Product changing the scope where this key moment could (and should) be tested everywhere via an A/B test.

However, I was ready for this scope change because I was already thinking about how we could easily tweak these designs to scale our approach.

What happened?

Results . Reflection

RESULTS

Platform Specific Impact
The A/B test revealed a significant platform divergence—the same designs performed differently across web and mobile.
Impact Summary
Responsive Web
No impact
Mobile app
+9.5% deposit increase
Why the Platform divergence?
During retros, we identified the key difference:
Web Finding
More "noncommittal users"
New users come through different marketing channels with lower intent. The friction wasn't the primary barrier—commitment was.
Mobile Finding
Higher baseline engagement
When you download an app, you're more likely to engage with it. This was especially true for Keen since it's a niche product. These users had intent—friction was the barrier.
THE INSIGHT
Cognitive load reduction has the biggest impact when users already have intent. Mobile app users had intent; web users often didn't. Different problems require different solutions.

Reflections

Strategic UX Leadership Isn't Just About Craft Excellence
Strategy-First Mindset
By framing this as a business problem (mobile-web conversion gap threatening competitive position) rather than a UI problem (cluttered interface), I gained stakeholder buy-in for thorough iteration and expanded scope when data showed opportunity beyond just new users.
Systems Thinking Over Feature Thinking
Diagnosing the root cause—mental model mismatch between where users were psychologically (trust formation) vs. where the system treated them (committed buyers)—revealed leverage points that surface-level UI polish would have missed.
Sections