Free lesson experiment
+13%
Relative premium conversion uplift
3 free lessons1.15%
2 free lessons1.3%
Tighter free access increased conversion from 1.15% to 1.3%.
06 of 06 · Commercial
Improving premium conversion through experimentation.
I tested free-access limits and paywall hierarchy as focused experiments to balance customer value with sustainable subscription growth.
At a glance
- Problem
- The product needed to let new learners experience enough value to understand Kaizen while maintaining a clear reason to subscribe.
- Key insight
- Conversion was sensitive not only to the amount of free access, but also to how subscription choices were framed and prioritised.
- Decision
- Test lesson limits and paywall hierarchy as focused experiments rather than redesigning the entire subscription journey.
- Result
- Reducing daily free lessons from three to two produced an approximately 13% relative premium-conversion uplift, while repositioning the recommended plan increased six-month subscriptions by more than 20%.
01 Commercial tension
Conversion and subscriptionexperiments.
Kaizen needed to let new learners experience enough value to understand the product while maintaining a clear reason to subscribe. I tested lesson limits and paywall hierarchy as focused experiments rather than redesigning the entire journey.
02 Commercial monitoring
Connecting experiments to performance
I monitored subscription volume, churn, plan length, acquisition and net customer growth so product changes could be assessed against commercial outcomes.
See how the same analytics infrastructure supported retention.
Dashboard snapshot · Aug–Nov 2019
361
Net customers
479
New subscriptions
3,900
First opens
Subscription type
- New479
- Repeat128
- Churned118
Plan length
- 1 month69.7%
- 3 month18.2%
- 6 month8.6%
- 12 month3.5%
Acquisition spend
Facebook£7,161
Apple£2,611
Free lesson experiment
- Hypothesis
- A tighter daily free-lesson limit would create a clearer reason to subscribe without withholding all value.
- Variants
- Variant A: three free lessons per day. Variant B: two free lessons per day.
- Result
- Conversion changed from 1.15% to 1.3% — approximately a 13% relative premium-conversion uplift.
- Implication
- This was not about adding friction for the sake of conversion. It was about finding the right balance between giving users enough value to understand the product and creating a clear reason to subscribe.
Paywall hierarchy experiment
- Hypothesis
- Clearer framing of the recommended plan would change plan selection without a full paywall redesign.
- Variants
- Most Popular label on three-month plan versus six-month plan, with one, three, six and twelve-month options.
- Result
- Six-month subscriptions increased by more than 20% when the recommended label moved to the six-month plan.
- Implication
- Choice architecture mattered as much as price — clearer framing changed decision-making.
Explored for a future phase
03 Additional exploration
Reducing sign-up friction with social sign-in
External research indicated that offering familiar sign-in options such as Apple, Google and Facebook could improve conversion by around 20%, with some implementations reporting uplifts as high as 40%. Based on that evidence, I recommended adding social sign-in to Kaizen rather than relying only on the existing registration route.
The founders were concerned that third-party authentication could expose customer information to providers such as Google and Facebook. I argued that excluding these familiar options risked reducing sign-up and sign-in completion, while a data-minimised implementation could request only the information required to authenticate an account and retain email registration as an alternative.
The recommendation was not implemented during this phase, so the external figures are presented as evidence supporting the decision—not as measured Kaizen impact.
- Opportunity
- Potentially improve sign-up and sign-in completion through familiar authentication options.
- Trade-off
- Balance conversion opportunity with privacy, permissions and third-party data concerns.
04 Verified outcomes
Most Popular label test
+20%
Six-month plan selection
Before
3 mo
Most Popular
After
6 mo
Most Popular
Moving recommendation emphasis to the six-month plan increased selection by more than 20%.
05 Results and learning
Focused experiments over full redesign.
Commercial optimisation did not require withholding all value or redesigning the complete journey. Focused changes to access and choice architecture created a clearer path from experiencing the product to understanding why a subscription was worthwhile.
The two verified experiments showed that conversion responded to both how much was free and how plans were framed — product thinking applied to commercial decisions.
This was the final story in the Kaizen Languages case study. Return to the overview to see how these product areas connected, or explore more work across the portfolio.