ZG

Course Vaults

Conversion is a timing problem

Role

Product Designer

Frontend & Backend Engineer

Platform

Mobile (B2C) App

Skills

Product Design, Monetization, Onboarding, Backend DB Infrastructure

New users built their golf history quickly, but usage dropped after the first session. I moved the subscription prompt into that initial high-intent experience: users could save 20 courses for free, then encountered the paywall when they tried to add course 21. Early free-to-paid conversion increased from 4.5% to 5.85%.

Before
4.5%
After
5.85%
Improvement
+1.35

percentage points

+30% relative

What is Course Vaults

Course Vaults is a web and mobile product where golfers track, rate, and revisit the courses they have played. Each rating also contributes to a shared, player-driven view of golf courses around the world.

It is a personal golf diary and a community rating system, similar in spirit to Letterboxd, but for golf.

Context

The contribution-focused onboarding worked: new users added courses quickly and began building a meaningful personal history during their first session.

But that early activity did not translate into sustained usage or paid conversion. Engagement peaked immediately after signup and then declined sharply. By the time users encountered a reason to subscribe, much of their initial intent had already disappeared.

Usage after download

Can we monetize this?Drop-offDay 1Week 1Week 2Month 1
The shape of early usage: a fast open, then a fast drop. The paywall asks during the spike.

The problem

Free-to-paid conversion measures the share of free users who eventually subscribe. Course Vaults was waiting too long to communicate the value of its paid plan.

The strongest expression of intent happened while users were actively adding courses they had already played. That high-intent moment was not connected to the subscription experience.

The thesis

Our hypothesis was that conversion depended partly on when we asked, not only what we charged.

If the subscription prompt appeared while users were actively building a golf history, the value of continuing would be more immediate and concrete.

The decision

We allowed users to save 20 courses for free. When they attempted to save course 21, Course Vaults presented the subscription paywall.

Twenty courses gave users enough room to understand the product and create a meaningful diary. The next save created a clear decision point: subscribe to continue expanding the history, or keep the existing free collection.

  1. 1

    Add courses

  2. 2

    Reach 20 free saves

  3. 3

    Attempt course 21

    Subscription prompt

The tradeoff

This decision introduced a real marketplace tradeoff. Course ratings improve the shared dataset, so limiting free contributions could reduce the number of ratings added to the platform.

By this stage, the product had accumulated enough initial course data to test monetization. We chose to prioritize learning about paid demand, while monitoring the effect on contribution volume.

Result

Early free-to-paid conversion increased from 4.5% to 5.85%, an improvement of 1.35 percentage points, or 30% relative to the previous rate.

The result supports the timing hypothesis, but it should be treated as directional until the behavior is observed across a larger population and longer measurement period.

Takeaways

  • Strong onboarding does not automatically produce retention or paid conversion.
  • Subscription prompts are easier to understand when they appear at a moment of demonstrated intent.
  • Report both absolute and relative conversion changes; 4.5% to 5.85% is +1.35 percentage points and +30% relative.
  • Monetizing contribution creates a tradeoff between subscription revenue and growth of the shared dataset.
  • More subscriptions do not necessarily mean proportionally more recurring revenue; pricing and plan mix must be evaluated separately.

Next steps

The next question is not only whether 20 is the right threshold. It is whether the additional subscribers retain, renew, and generate enough recurring revenue to justify the contribution limit.

The next tests should isolate one variable at a time:

  1. Keep pricing constant and compare course limits such as 20 versus 30.
  2. Keep the winning threshold constant and compare a simplified monthly-and-yearly plan structure.
  3. Measure ARR per eligible user or checkout visitor, not subscription count alone.
  4. Compare retention, renewal, refunds, and contribution volume across cohorts.
  5. Test ways to bring users back after their initial golf history is complete.

The long-term goal is not simply to show the paywall earlier. It is to find the point where user value, contribution growth, retention, and recurring revenue reinforce one another.

2026