{"id":3654,"date":"2025-08-21T18:23:27","date_gmt":"2025-08-21T18:23:27","guid":{"rendered":"https:\/\/persona.ly\/blog\/?p=3654"},"modified":"2025-09-02T11:59:12","modified_gmt":"2025-09-02T11:59:12","slug":"how-we-outperformed-roas-d30-and-re-engaged-high-value-audience-for-golf-king","status":"publish","type":"post","link":"https:\/\/persona.ly\/blog\/2025\/08\/how-we-outperformed-roas-d30-and-re-engaged-high-value-audience-for-golf-king\/","title":{"rendered":"How We Outperformed ROAS D30 and Re-engaged High-value Audience for Golf King – World Tour"},"content":{"rendered":"\n

Download the pdf version of the case study<\/a><\/em><\/p>\n\n\n\n

\n
\n
\"GolfKing-app-icon-re-engagement-case-study\"<\/a><\/figure>\n<\/div>\n\n\n\n
\n

We partnered with RisingWings<\/a> to re-engage lapsed users of their Golf King<\/a> app on Android in the US<\/strong>. RisingWings is a top-grossing Korean game developer specializing in multiplayer simulation games enjoyed by hundreds of millions of users worldwide. Golf King is a real-time 1v1 golf duel game featuring customizable characters, clubs, realistic courses, and weekly leagues.<\/p>\n<\/div>\n<\/div>\n\n\n\n

<\/div>\n\n\n\n

Results at a Glance<\/h2>\n\n\n\n
\n
\n

35%<\/strong><\/p>\n\n\n\n

Uplift in D30 ROAS within 3 months<\/p>\n<\/div>\n\n\n\n

\n

27%<\/strong><\/p>\n\n\n\n

Decrease in D30 Unique Purchase cost by month 3<\/p>\n<\/div>\n\n\n\n

\n

16%<\/strong><\/p>\n\n\n\n

Growth in Average Number of Purchases per Paying User by month 3<\/p>\n<\/div>\n<\/div>\n\n\n\n

<\/div>\n\n\n\n

Campaign Goal<\/h2>\n\n\n\n

The goal of the campaign was to bring churned users back to the app to make either a new in-app purchase (Payers segment) or their first in-app purchase (Non-payers segment), with D30 ROAS being the KPI.<\/p>\n\n\n\n

<\/div>\n\n\n\n
\n
\"\"<\/a><\/figure>\n\n\n\n
<\/div>\n\n\n\n

\u201cPersona.ly has been an incredible partner in keeping our campaigns running smoothly. The team is reliable, truly committed to our goals, and takes the time to understand our game\u2019s needs, building smart, performance-driven strategies around them.<\/p>\n\n\n\n

We\u2019ve seen consistent improvements in results, and their knack for handling both technical and budget challenges while keeping our priorities front and center makes them a partner we can always count on. We\u2019re excited to keep working together and explore new ways to drive even more value.\u201d<\/p>\n\n\n\n

<\/p>\nYee Seul Hwang \ud669\uc774\uc2ac – UA Team Lead at Rising Wings<\/strong><\/cite><\/blockquote>\n\n\n\n

<\/div>\n\n\n\n

Programmatic Re-engagement Process <\/h2>\n\n\n\n

At the beginning of our collaboration, using first-party data received from the client and our proprietary Live Audiences feature <\/a><\/strong>(audience segmentation engine available to our clients at no cost), we generated two audience segments:
– Payers:<\/strong> Users who had previously made in-app purchases and stopped using the app.
– Non-payers (NPUs): <\/strong>Users who played Golf King in the past, didn\u2019t make any in-app purchases, and churned later.<\/p>\n\n\n\n

Having access to first-party data allowed us to generate the segments based on users\u2019 purchase intent <\/strong>and refine our targeting strategy to drive better performance across both audiences.<\/p>\n\n\n\n

Our extensive experience in the gaming vertical allowed us to reach and outperform the customer\u2019s ROAS-based KPI<\/strong> in the first month of running the campaign. As we received more positive signals on which users are more likely to re-engage (re-install or re-open the app) and make in-app purchases, the learning progressed, and our ML-driven bidder was able to target those audience segments more precisely.
<\/p>\n\n\n\n

\n
\n
\"Persona.lys-Dynamic-Audiences-feature\"<\/a><\/figure>\n<\/div>\n\n\n\n
\n

Our programmatic ML-driven DSP is integrated exclusively with tier-1 ad exchanges and processes over 4 million ad requests per second. This allows us to quickly gather data for our bidder to optimize the campaigns and bring high-quality traffic at scale.<\/p>\n<\/div>\n<\/div>\n\n\n\n

By strategically allocating budget toward the most prominent audience segments with the highest intent to make in-app purchases, we continued optimizing and improving the performance of the campaign.

The steady growth in D30 ROAS<\/strong>, coupled with a decrease in D30 Unique Purchase Cost and an increase in the average number of purchases per paying user, demonstrated the effectiveness of our segmentation and optimization strategy. This success quickly built confidence in our ability to deliver scalable results, fostering a lasting collaboration with RisingWings<\/strong>.<\/p>\n\n\n\n

<\/div>\n\n\n\n

Campaign Results<\/h2>\n\n\n\n

D30 ROAS Growth Trend  <\/h3>\n\n\n\n
\n
\n

From the outset of the campaign with RisingWings, we quickly managed to reach and outperform the D30 ROAS KPI. As the learning progressed, our programmatic bidder was able to distinguish between users who would re-engage and make in-app purchases (in both Payers and Non-payers segments) and those who wouldn\u2019t.<\/p>\n\n\n\n

We continued optimizing the campaign, yielding performance 35% above the KPI by month 3.<\/strong><\/p>\n<\/div>\n\n\n\n

\n
\"\"<\/a><\/figure><\/div><\/div>\n<\/div>\n\n\n\n

Decrease in Unique Purchase Cost and Growth in Average Number of Purchases per Paying User<\/h3>\n\n\n\n

As soon as we identified users with a higher intent to make in-app purchases in both segments (unique repurchases for the payers segment and first-time deposits (FTD) for the non-payers segment), the cost of unique purchases started to decrease.
By optimizing toward the ROAS-based KPI, our programmatic bidder leveraged the growing dataset to target user cohorts more likely to make multiple purchases. As a result, the average number of purchases per paying user increased.<\/p>\n\n\n\n

\n
\n
\"\"<\/a><\/figure>\n<\/div>\n\n\n\n
\n

By continuously optimizing the campaign towards the most prominent audience cohorts, by month 3 we achieved: <\/p>\n\n\n\n