Where AI Marketing Begins?

Who to Target, What to Say, How to Say

Game marketing usually starts in earnest when localization begins. This is when the business team creates launch plans, aligns internally, presents to leadership, and secures budget — all prerequisites before any campaign can be executed.

When it is time to design the campaign, many teams still ask the same question: Where do we start?

Should we aim for a global simultaneous launch?

Which user acquisition channels make the most sense?

Who will create the creatives, and how will we operate the campaign?

As the list grows, these discussions lead to the most fundamental question:

ℹ️ Why “Everyone” Is No Longer the Target

Targeting broad groups like “males aged 18–24 who like games” is no longer effective.

Games now compete in the broader digital entertainment space, not only with other games but also with Netflix, YouTube, and TikTok. In this environment, simply making a good game is not enough. Precisely defining your audience and building strategy around them has become a core requirement for success.

This is where AI-based persona strategies come in.

ℹ️ AI-Powered Segmentation: Defining Users Before You Meet Them

AI can analyze user data to automatically generate detailed personas. By combining behavioral data such as play logs, purchase history, churn timing, and social interactions, it identifies users with similar patterns and groups them together. This process is known as clustering

For Example:

• Users who play for 30 minutes on weekday evenings but never spend

• Users who log in only during events and make short bursts of high spending

AI goes beyond demographics such as gender, age, and location. It analyzes emotional expression, communication style, and reasons for churn to infer the group’s characteristics and response patterns. Here’s how it works:

� Jean Kim (Female, 35)

Logs in 7–8 AM and after 11 PM

Prefers collection of content over competition

Has app store purchase experience

Avoid community forums but responds to push notifications

→ A “short-session immersion” player who relieves stress with mobile puzzle games before work and at night

ℹ️ No Data Yet? How to Start AI Personas Before Launch

For new titles without internal player data, building AI personas may feel difficult. But with external data and predictive AI tools, you can start right away. For example:

Analyze reviews, community discussions, and play records from similar games

Use early signals from pre-registration or beta users, such as click rates and time spent

Explore potential user segments with tools like Appier or Delve AI

In today’s cookieless environment, relying on external tracking alone is no longer effective. Instead, genre-specific behavioral data, community reactions, in-game signals, and AI-based segmentation have become even more valuable.

AI helps predict the behavior of users you have not yet met. When your own data is limited, the strategy begins by leveraging data that already exists.

ℹ️ Validating AI Personas Against Real Players

AI-generated personas must always be validated against reality. In practice, two approaches are common:

① Behavioral validation

for instance, if AI predicts strong interest in the in-game shop, verify this with actual visit logs

② User feedback validation

compare AI-inferred traits with interviews or survey responses

Through this validation, personas become more accurate and gradually evolve into actionable strategic units.

ℹ️ Testing Strategies With A/B Experiments

AI-generated personas are no longer mere hypotheses; they are now testable strategic units that can be validated through A/B testing by quantifying responses. For example:

One group sees a “competition-focused” message

Another group sees a “story immersion” message,

Compare each response (such as CTR, conversion rate, etc.).

Through this process, you can identify with data which persona responds to which message.

This shifts the question from simply asking “What worked?” to understanding “Who did it work for, and what exactly resonated with them?”

ℹ️AI That Remembers Experiments: Automating Optimization

Even after A/B tests conclude, AI continues to learn. It applies successful patterns to future campaigns, prioritizes combinations that deliver better outcomes, and automatically adjusts persona definitions as user behavior evolves.

In this way, AI becomes a continuously improving strategic partner that remembers, adapts, and optimizes.

ℹ️ The Marketer’s New Role: Strategy Architect

AI-based persona strategies enable marketers to design who to target, what message to deliver, and how to deliver it — not based on intuition but grounded in data and experiment results.

In a cookieless environment, funnel design can no longer rely solely on external advertising platforms. Instead, marketers need to build strategies that leverage first-party data effectively with AI. This redefines the marketer’s role as:

Leader – setting the direction for marketing and user data strategy

Curator – providing AI with the right context for analysis

Designer – refining strategies through continuous experimentation

This is the new role marketers must embrace in today’s game industry.

In the next section, I will share practical strategies for running AI-driven marketing campaigns, along with real use cases from the gaming sector

※ Disclaimer: This content reflects the author’s personal views and includes only publicly available examples. It does not represent the official position of any company mentioned