Targeting

What is Targeting?

The core logic of Targeting

• Data foundation : Relying on multi-dimensional data such as user device information (such as model, operating system), geographical location, browsing behavior, interest tags, etc., to build a profile.
• Algorithm-driven： Utilize machine learning models to predict the probability of user response to advertisements, dynamically optimizing delivery strategies.
• Real-time feedback： Adjust targeting parameters based on real-time metrics such as click-through rate and conversion rate to form a closed-loop optimization.

Typical application scenarios of targeting

1. User Profile Targeting

• Demographic attributes： Basic labels such as age, gender, occupation, income level, etc.
• Interests and Preferences： Interest classification inferred from APP usage records, search keywords, and content interaction behaviors.
• Consumption capacity： Determine the user's willingness to pay based on their historical purchase records or credit card binding status.

2. Behavioral trajectory targeting

• Search keyword targeting： Target users who have recently searched for "travel guides" to push flight/hotel ads.
• In-app behavior targeting ： Conduct cross-platform retargeting ads for users who have browsed sneakers within the e-commerce app but have not placed an order.
• Location-based targeting (Geofencing): Pushing discount information of nearby stores to users within a 1-kilometer radius of the business district.

3. Contextual targeting

• Time-sensitive targeting： During breakfast hours, push coffee coupons; in the evening, promote video membership sales.
• Device environment orientation : Decide whether to deliver high-traffic video ads based on the network environment (WiFi/4G).
• Content relevance targeting： Place sports equipment ads on fitness content pages.

4. Advanced targeting technology

• Similar audiences： Expand the similar potential customer group through seed user characteristics.
• Cross-device ID mapping： Identify the behavioral correlation of the same user across different terminals (mobile phone/tablet/PC).
• Predictive targeting： Based on historical data, predict the user's shopping demand categories for the next 3 days.

The core metric for evaluating the effectiveness of targeting.

1. User Reach Dimension

• Coverage (Reach Rate): The proportion of users actually reached within the target audience.
• Frequency control (Frequency Capping): The maximum number of times a single user can receive the same advertisement.
• Invalid traffic rate (Invalid Traffic Rate): The proportion of clicks from bots or anomalies.

2. Interaction Efficiency Dimension

• Click-through rate (CTR): The proportion of click-through actions among ad impressions.
• Effective viewing rate (Viewable Rate): The proportion of video ads actually viewed for ≥2 seconds.
• Interaction depth ：Behavior data such as the duration of user stay after clicking and the depth of page scrolling.

3. Conversion Effect Dimension

• Conversion rate (CVR): The proportion of users who complete registration/purchase and other target actions among those who click.
• Cost per conversion (CPA): The ratio of total advertising costs to the number of conversions.
• Return on Advertising Investment (ROAS): Direct revenue generated for every 1 yuan spent on advertising.

4. Long-term value dimension

• User retention rate： The proportion of users who remain active 7 days/30 days after being acquired through advertising.
• Lifetime value (LTV): The total revenue generated by a user throughout their entire usage period.
• Brand search index ：The increase in search volume for related brand keywords after the advertisement is released.

The Evolution Trend of Targeting Technology

1. Data compliance:

2. Model intelligence:

3. Cross-platform integration:

4. Experience personalization: