Optimizing Membership Upgrade Rates with a 7-Day Free Trial
A data-driven approach to increasing user conversions and financial well-being.
Duration
15-day experiment
July 11, 2024 - July 26, 2024
Industry
Fintech (Savings and Credit Building)
Resources/Tools
Figma, Count, Miro and Webflow
Role
Product Designer
User Research, Interaction, Visual design, Prototyping & Testing
Background
Loqbox is a UK-based fintech company specializing in helping individuals build credit and save money. Through its unique approach, Loqbox allows users to save money in a way that also builds their credit score, empowering them to access better financial products in the future.
Loqbox offers two membership options: Full Membership and Lite Membership. Full members enjoy a wider range of benefits, including faster credit score growth and enhanced customer support. Learn more about Loqbox here
Approach
Proposed solution
To address the challenge of increasing upgrade rates without sacrificing revenue, proposed testing a 7-day free trial with three different price points (£2.50, £2.99, £3) and removing the Lite membership option. This approach aimed to simplify the user decision-making process while testing the impact of pricing on conversion rates.
Tools used
Figma: For wireframes and high-fidelity designs.
Count Dashboard: For data analysis and visualization.
Supporting research
Key Insights from Downgrade Research
Experiment Design
Implementation
Redesigned the Member First sign-up journey to remove the Lite membership option and introduce a 7-day free trial. Three price points (£2.50, £2.99, £3) were tested to determine the optimal pricing strategy while ensuring the design aligned with Loqbox’s mission of transparency and financial empowerment.
A/B Testing Setup
Ran a 15-day A/B test with four groups:
Control Group: No free trial, Lite option available, £2.50 price.
Variant A: 7-day free trial, no Lite option, £2.50 price.
Variant B: 7-day free trial, no Lite option, £2.99 price.
Variant C: 7-day free trial, no Lite option, £3.00 price.
The experiment was implemented at 100% of Member First traffic i.e., 25% per group, and monitored upgrade rates, payment retention, and downgrade rates.
Explore the Figma here
Timeline and Key Details
Go-Live Date: April 4, 2024
End Date: April 18, 2024
Total Duration: 14 Days
Traffic: 100% of users in the downgrade journey participated (50% to the Control & 50% to the Variant)
Journey: Downgrade journey
Challenges
1. Balancing Revenue Loss and Upgrade Rates: One of the main challenges was balancing the potential revenue loss from offering a free trial with the goal of increasing upgrade rates. In MX-115, the 4-week free trial increased upgrade rates but resulted in a significant loss of initial revenue. We needed to determine if a shorter 7-day free trial could achieve similar results without sacrificing too much revenue.
2. Price Sensitivity: In MX-113, increasing the price from £2.50 to £3 without a free trial reduced upgrade rates by 3%. We were unsure if users would respond differently to higher price points when paired with a free trial. This uncertainty required careful testing and analysis.
3. Downgrade Rates: In this experiment, we observed slightly higher downgrade rates for Variant users compared to the Control group. This was unexpected, as previous experiments (MX-115) showed better retention for free trial users. We are closely monitoring this to understand if it’s a short-term effect or a long-term trend.
4. Interaction Between Free Trial and Price: We identified a potential interaction effect between free trials and price points. Specifically, £3 underperformed without a free trial but outperformed with one. This required further investigation, and we are planning a 2×2 factorial experiment to verify this interaction.
Impact & Results
Revenue Impact
Winning Variant (Variant C): The 7-day free trial with a £3 price point performed the best, achieving:
Upgrade Rate: 45.80% (vs. 37.66% for the Control).
Payment Retention Rate: 72.15% (vs. 70.59% for the Control).
Business Outcome: Variant C was implemented across the Loqbox web app, driving better user engagement and revenue growth. Users were willing to pay the higher price (£3) after experiencing the free trial, indicating strong value perception.
Key Metrics
1. Upgrade Rate
Control Group: 37.66%
Variant A (£2.50): 39.45% (slightly higher than Control).
Variant B (£2.99): 43.78% (noticeably higher than Control).
Variant C (£3): 45.80% (highest upgrade rate, significantly outperforming Control).
Insight: Variant C’s higher price point (£3) did not deter users; instead, it led to the highest upgrade rate, suggesting that users perceived greater value in the offering.
2. Payment Retention
Control Group: 70.59%
Variant A: 75.00% (highest retention, but with a lower upgrade rate).
Variant B: 69.17% (slightly lower than Control).
Variant C: 72.15% (strong retention, second only to Variant A).
Insight: Variant C strikes a good balance between upgrade rate and payment retention, making it the most effective variant overall.
3. Downgrade Rates
Variant users had slightly higher downgrade rates, but this did not significantly impact overall retention.
Insight: The higher downgrade rates may be due to users reassessing their needs after the free trial, but the overall retention metrics suggest this is not a major concern.
Statistical Significance
Variant C vs. Control: p = 0.011 (significant)
Interpretation: The p-value of 0.011 is less than the 0.05 threshold, indicating that the difference in upgrade rates between Variant C and the Control is statistically significant.
Conclusion: Variant C’s higher upgrade rate is unlikely due to random chance, and the £3 price point is a clear winner.
Variant B vs. Control: p = 0.078 (borderline significant)
Interpretation: The p-value of 0.078 is slightly above the 0.05 threshold, suggesting a borderline significant difference.
Conclusion: Variant B (£2.99) shows potential, but the evidence is not strong enough to be conclusive. It may be worth further investigation or testing with a larger sample size.
Variant A vs. Variant C: p = 0.054 (borderline significant)
Interpretation: The p-value of 0.054 is close to the 0.05 threshold, indicating a borderline significant difference.
Conclusion: Variant C (£3) likely outperforms Variant A (£2.50), but the evidence is not definitive. This suggests that the £3 price point may be more effective, but further testing could confirm this.
Conclusion
By combining a 7-day free trial with a higher price point (£3) and removing the Lite option, we were able to significantly increase upgrade rates (45.80%) while minimizing revenue loss. This project demonstrates the power of data-driven design in solving complex business problems and aligns with Loqbox’s mission of empowering users to achieve financial stability. The winning variant is now in production across the Loqbox web app, driving better user engagement and business outcomes.