How a Prominent Gaming Company Drives Profitable Growth by Predicting the Sustainability of New Players Early in Their Lifecycle With 95% Accuracy

With high, double-digit year-over-year churn, an online sports betting and iGaming platform had to build a new player base every year. The platform realized that they did not know how to attract or retain the right players. Their production and transaction data alone was not enough to accurately identify their highest-value players early in their lifecycle.  


Acquiring a new player base every year is expensive; it’s much more cost effective to retain the best players for a longer time. The best way to do this is to target sustainable players and leads that will provide a high lifetime value (LTV) far beyond their acquisition cost.

They faced four major challenges:

  • Very high YOY churn
  • Limited understanding of the profile of high-value players at the aggregate level
  • Low accuracy in identifying high-value players at the individual level
  • Unprofitable players taking advantage of promos

The gaming app had originally defined high-value players by their number of deposits and frequency of game play within the user’s first week. In attempts to win back churned players, they offered blanket promotions — which sometimes made the reactivation costs 20 times more expensive than the original acquisition. The brand also concentrated most of their marketing efforts on Facebook, believing it was the source of the most conversions and high value players.

In a bid to woo new players, the game offered a $100 free deposit amount, and quickly noticed that these users would download the app, gamble the free money, cash out, and delete the app.

This one-size-fits-all approach didn’t work because the company had zero confidence in their ability to identify their high value players. They incorrectly defined high-value users as those with the most deposits and game interaction in the first week with the app.

The company set a goal of increasing their percentage of high-value players from 30% of their player base to 50%. They calculated reaching this goal would net a cash increase of $1.6 million a year.

Results with Ocurate

The gaming platform initially asked Ocurate to identify which players would lose the most, thinking that would be the key to more revenue. However, Ocurate’s data simulation of the platform’s four-year historical player data instead showed that players who lose and win over a longer period of time not only stay with the game longer, they also spend three times more money than lower value, short-term players.

Ocurate’s deep learning framework to predict LTV at the individual level increased the brand’s LTV prediction accuracy by 35%.

Highly accurate LTV predictions enabled the brand to:
Correctly identify 25% more high-value players in the game’s existing database
Make smarter investments to acquire and retain sustainable players

Powered by these insights, their General Manager calculated a potential gross profit increase of 19.6%. 

Because LTV tends to be calculated in retrospect and/or in the aggregate, it can be difficult to predict LTV for individual players early in their lifecycle. 

This is the Ocurate advantage: Using a dataset of in-depth behavioral and attitudinal profiles for each adult in the United States combined with the brand’s customer data (customer ID and production, telemetry, and purchase data), Ocurate created a detailed profile of the brand’s ideal, high-value players. This profile identified where these players spend their time online, their personality, financial status, interests, and behaviors. 

At the individual level, Ocurate predicts the degree of sustainability within 3 days of play at 81% accuracy. Accuracy increases to 95% within a player's first three weeks on the platform.

Ocurate's Sustainable Probability
"Ocurate has demonstrated that it can predict lifetime value for our customers early in their customer journey with proven accuracy. Implementing Ocurate will help us reduce churn and achieve sustainable growth, increasing the number of high-value customers." - General Manager

With this information, the brand and their marketing agency took concrete, data-based actions to improve their acquisition and retention tactics, including:

  • Providing dynamic commissions to affiliate partners depending on player sustainability
  • Creating targeted winback campaigns with promos based on LTV and changes to a player’s lifestyle (i.e. new home purchase, relocation, etc.)
  • Remove players “gaming” the system from the platform

After using Ocurate for one week, the brand realized that the way they calculated LTV was wrong; they were using antiquated models to analyze data in their CRM. They realized their belief that Facebook was their highest performing channel was incorrect — in fact, Facebook brought them customers of the lowest value, and only 10% of customers coming from Facebook were profitable.

Acting off of Ocurate’s insights, the brand reduced their investment in Facebook and lowered their promo for customers acquired through Facebook. Now, they are prioritizing investments in more profitable marketing partners with the goal of improving their LTV:CAC ratio. 

Ocurate created over a 35% more LTV customers

Ocurate’s unique benefits

Superior data

We took five years of R&D to build a future-proof database that maps every adult Americans’ foundational personality, behavioral, and attitudinal traits with explicit consent.

Better AI

PhD-level data scientists from Stanford, Wharton, Columbia, and Cornell built our deep machine learning framework, which predicts LTV with more than 90% accuracy.

Seamless integration

Insights are tailored to industry, business, and player base by examining historical player data (like transaction history, production data, and player ID). We integrate real-time predictions into your existing workflows via API — no matter where or how your customer data is stored.

Ocurate will soon release open-source software for lifetime value modeling. Want to get early access? Sign up below to join the waitlist.