How PropGPT increased proceeds per user by almost half with a simple price increase test

Intro

PropGPT is a sports betting assistant app, where users pay for valuable, data-driven insights on player props and game outcomes. In this competitive niche, maximizing the revenue captured from each user is crucial for profitability.

The PropGPT team wanted to validate a fundamental question: Would increasing the subscription boost revenue from each user in the long term?

The challenge

PropGPT needed to determine their price elasticity:

• Revenue uncertainty. The team suspected they were leaving money on the table with their current weekly rate, so wanted to test a higher price without risking their user base.

• Hypothesis. They believed their value was high enough that a 43% price increase wouldn't deter conversion, but would increase revenue from every new customer.

• Risk mitigation. Before deploying to their whole base, they needed an A/B testing environment to split traffic, test their hypothesis, and measure Proceeds Per User (PPU).

The solution

PropGPT executed a simple, but high-stakes, price point A/B test:

• Control cohort. Saw the existing subscription price of $6.99.

• Variant cohort. Saw the proposed new price of $9.99, with identical images and copy.

They A/B tested two distinct cohorts of users simultaneously, allowing them to measure the financial impact of the higher price on conversion and subsequent revenue.

The results

The experiment provided a definitive answer to the price elasticity question and a massive gain:

• PPU soared. The variant cohort that saw the $9.99 price point delivered a 47.5% increase in PPU.

• Hypothesis validated. Conversion rate erosion at the higher price was minimal, showing the trade-off was more than worth it.

• Value maximized. By increasing price, PropGPT successfully captured the maximum possible value from their audience.