General Tracking - Usage Recommendations

1. Macro Ad Date

💡In the Playturbo event analysis dashboard, use the trend analysis module to compare data across various macro dimensions.

Dimension	Example	Recommendations for Use
Ad Network	Mintegral/Unity	Compare the performance of creatives across different channels. Identify creative preferences within different channels.
Target Area	CN/US/JP	Compare the performance differences of creatives in different areas. Which creatives work best for each area.
Mobile OS	IOS/Android	Compare creative effectiveness on different mobile os.
Ad Type	IV/RV	Different ad types have varying display logic, so compare them separately. Compare creative effectiveness for different ad types. Investigate if consistently high or low metrics are due to a specific ad type's characteristics.
Project Name/ Associated Product	-	Product's current distribution status and variations in data between different products. Identify projects with strong overall distribution performance for scaling and those with lower-than-expected data for improvement.

For dive deeper into, refer to the following approach to analyze across and individual creatives.

2. Creative Basic Tracking

💡Comparing tracking can help pinpoint underperforming creative and identify the step for improvement.

💡The commonly used tracking types can be categorized into conversion metrics, interaction metrics, and loss metrics.

STEP1：Check if Creatives Lead to Desired Redirection through Conversion Metrics

Metric	Definition	Data Interpretation
Redirect Rate	=Redirect / PL View	High: Overall performance of the creative is good, users are more likely to trigger active/passive clicks; The number of interactive mechanisms is sufficient and working correctly; Further optimization efforts can focus on improving interaction rates, installation rates, and other deep metrics. Low: Insufficient click guidance or ineffective click mechanisms.
Redirect Rate during Gameplay	=Redirect during Gameplay / PL View	The metrics on the left can be used to confirm the effectiveness of the redirect mechanism. If the rate is high at a certain step, it indicates that the redirect mechanism at that stage is effective. If a redirect mechanism is set for a specific step but the conversion rate is low, it is necessary to investigate whether the redirect function is working properly.
Redirect Rate after Gameplay	=Redirect after Gameplay / PL View	The metrics on the left can be used to confirm the effectiveness of the redirect mechanism. If the rate is high at a certain step, it indicates that the redirect mechanism at that stage is effective. If a redirect mechanism is set for a specific step but the conversion rate is low, it is necessary to investigate whether the redirect function is working properly.

Technical Logic Expansion Explanation:

1. In the event analysis, all conversion metrics are deduplicated based on unique requests and are not double-counted.

2. All conversion metrics are only counted for the first occurrence. For example, redirection followed by a return to continue, and then another redirection afterward will only be counted as one redirection during gameplay.

3. All "PL View" in this document are counted based on the number of game starts.

STEP2：Evaluate User Actions at Key Points through Interaction Metrics

Metric	Definition	Data Interpretation
First Interaction Rate	=First Interaction / PL View	High: The opening scene is attractive, and the rules are easy to grasp, which can engage users in the first click. Low: A significant number of users lost at the first step, prioritize optimization of the opening guidance.
Game End Rate	=Game End / PL View	High game end rate or minimal difference from first interaction rate: Users have strong motivation, and low loss during gameplay, so the optimization should focus on the first interaction rate and conversion rate. Low game end rate: Users lose their goal or motivation during the process, leading to loss. It is recommended to optimize by strengthening anti-stalling guidance mechanisms, enhancing creative appeal, and reducing playing duration.
Replay Rate	=Replay / PL View	High replay rate indicates that users are more likely to click on "retry" when they fail: It indicates that the gameplay has some appeal and rekindle the user's desire to challenge again through inertia or provocative creative tactics. [Advanced Recommendation] During iterations, you can consider adding a redirect event to the "Retry" button when the retry attempts are exhausted. This will encourage users to convert when they are still in the flow.
TTE	=Average time from creative show to first interaction (excluding those who didn't interact)	Short: The creative can guide users to trigger their first interaction in a short time. Long: It may be due to insufficient user interaction motivation and a high understanding cost of the game. It is recommended to review and optimize the content of the creative.
Duration of Gameplay	=Average time from creative show to play end (excluding those who didn't complete play)	Duration of Gameplay can be used to measure the volume. By comparing the effectiveness of creatives with different volumes, you can summarize the most suitable duration.
Effective Interaction	=Number of interactive actions by users triggering game logic (Counting all instances, without deduplication)	It can be used to simulate and assess the total number of steps and process length for operations, helping evaluate the creative effectiveness. For example, if the game end rate and conversion rate are low but the effective interaction is high, you need to consider whether the process is too lengthy or there are too many levels, making it difficult for users to reach the conversion stage at the end.

Conversion Metric Reference（To be added）

Ad Type	RV	RV	RV	RV	RV	IV	IV	IV	IV	IV
Category/ Reference	First Interaction Rate	Game End Rate	Replay Rate	TTE	Duration of Gameplay	First Interaction Rate	Game End Rate	Replay Rate	TTE	Duration of Gameplay
Casual	-	-	-	-	-	-	-	-	-	-
Vertical-5	93%	35%	-	-	-	89%	28%	-	-	-
Mid-Core	-	-	-	-	-	-	-	-	-	-
Non-Game	-	-	-	-	-	-	-	-	-	-

STEP3：Locate the Step with High Loss through Loss Metrics

Metric	Definition	Data Interpretation
Lost Rate during Gameplay	=Loss during Gameplay / PL View	A low loss rate indicates that users are highly motivated to complete, and most users stay until the end. Optimization should be focused on improving the creative content and the inducement at the end to increase the chances of conversion and installation.
Loss Rate after Gameplay	=Loss after Gameplay / PL View	If the loss rate during gameplay is significantly higher than it after gameplay, it indicates that there is a more serious loss issue during gameplay.
Loss Rate before Interaction	=Loss before Interaction / PL View	By comparing the loss rates before and after interaction, attributions can be made. For example, if the loss rate is higher before interaction than after, it indicates that the creative's initial appeal needs improvement.
Loss Rate after Interaction	=Loss after Interaction / PL View	By comparing the loss rates before and after interaction, attributions can be made. For example, if the loss rate is higher before interaction than after, it indicates that the creative's initial appeal needs improvement.

Technical Logic Expansion Explanation:

1. The loss rate during gameplay is calculated only after first click.

2. Only creatives where users do not redirect but trigger game end are counted as loss after gameplay.

Loss Metric Reference：To be added

Other Metrics

Close Creative Rate：Close Creative/PL View

If the close rate is close to the game end rate, it indicates that majority of users are clicking the close button during the gameplay instead of completing it normally.It is recommended to optimize by strengthening anti-stalling guidance mechanisms, enhancing creative appeal, and reducing playing duration.

3. Scene Tracking and Custom Tracking

💡Analysis Premise:

1) Clearly define the analysis purpose as either A/B test or creative diagnosis, and choose the appropriate approach.

2) String scenes according to their actual logical sequence.

3) If you need to compare across creatives, ensure that their overall length and production methods are similar for higher comparability.

Metric	Definition	Data Interpretation
Scene N- Scene Arrival	=Number of times users arrived the current scene	If arrival count significantly decreases compared to the previous scene, it means that most users stopped in the previous one. So it's advisable to focus on optimizing the previous one.
Scene N- Rate of Redirection to Another Scene	=Redirect to Another Scene / Scene Arrival	If the rate is high, it indicates that users were motivated and most of them completed this scene. For the last scene, if there's no mechanism for replaying, this rate should be 0.
Scene N- Duration	=Average time of user stay in this scene (excluding those who have not navigated)	If the duration is too long, it indicates that this scene took a long time to play. If the duration doesn't meet expectations, it's advisable to investigate whether it's due to high difficulty, bugs, or other factors.
Scene N- Redirect Rate	=Scene Redirect / Scene Arrival	*If the scene doesn't have a redirection mechanism, you can ignore this metric. High: This scene effectively triggered the redirection mechanism, and almost all users reached this scene triggered redirection. Low: It may indicate insufficient guidance for redirection or that the redirection mechanism didn't work correctly.
Scene N- First Interaction Rate	=Scene First Interaction / Scene Arrival	High: The scene has a certain appeal, and users are willing to engage for the first time after entering the scene. Low: Many users lost after seeing the content of this scene. It is recommended to optimize the fun factor, feedback intensity, difficulty, etc..
Scene N- Lost Rate during Gameplay	=Scene Loss during Gameplay/ Scene Arrival during Gameplay	Low: Users have a strong motivation to complete the trial content of this scene. High: User attention and interest gradually decline at this stage. The reasons could be a dull process, lackluster feedback, performance issues, etc.
Scene N- Loss Rate before Interaction	=Scene Loss before Interaction / Scene Arrival before Interaction	By comparing the loss rate before and after interaction in a scene, you can attribute the situation. For example, if the loss rate is lower after interaction than before, it may be due to a poor post-interaction experience, causing players to lose the motivation to continue and leading to lost.
Scene N- Loss Rate after Interaction	=Scene Loss after Interaction / Scene Arrival after Interaction	By comparing the loss rate before and after interaction in a scene, you can attribute the situation. For example, if the loss rate is lower after interaction than before, it may be due to a poor post-interaction experience, causing players to lose the motivation to continue and leading to lost.

Technical Logic Expansion Explanation:

Redirection to another scene will be included in the replays but will not be included in the redirect to app store.

4.Case Studies

I. Purpose A：Creative Diagnosis

STEP 1: Identify the problem stage through scene tracking.

💡Once you've identified the key creatives and stages to focus on, you can use the creative analysis feature to pinpoint exact issues within each creative.

1. For scenes with poor tracking data (such as low arrival rate, low first interaction rate, and high loss rate), refer to the interpretation approach in the table above to pinpoint specific issues.

2. Target identified process, functional, and performance issues, determine the adjustment direction.

3. *If custom tracking is set, refine your investigation using Step 2.

STEP 2: Pinpoint loss points and refine optimization solutions through custom tracking.

💡Custom tracking is defined by the content creator and closely linked to gameplay details, using the format "actionXX". It pinpoints gameplay milestones more accurately than basic tracking.

Approach to Viewing Custom Tracking:

Determine the meaning of each 'action,' including some linear and non-linear trackings.

• Linear：It refers to trackings with fixed timing. When arranged sequentially, you can calculate the loss rate and arrival rate for each custom stage.

• Non-Linear：It refers to independent buttons not tied to a fixed process. They can be compared with related linear trackings to calculate metrics such as failure rate and specific item usage rate.

Case 1：Creative Diagnosis

Click to preview the case

1）Creative Basic Tracking

Tracking	Date	Data Interpretation（Orange = Key Conclusion）
Redirect Rate	40%	No clear issues, still room for improvement.
Redirect Rate during Gameplay	0%	No mid-game redirects, consider adding a download button for better conversion.
Redirect Rate after Gameplay	40%	No clear issues, still room for improvement.
First Interaction Rate	71%	Falls below the benchmark of an 80%+ first interaction rate for the casual category.
Game End Rate	52%	No clear issues, roughly 50% of users play through to the end.
Replay Rate	/	No replay mechanism, no analysis.
TTE	3s	Excellent performance. Users can quickly grasp gameplay and engage.
Duration	26s	Only 4 interactions, but the average duration is nearly half a minute, indicating waiting time between each action (e.g., watching animations). Consider reducing loss risk by accelerating or shortening animations as needed.
Effective Interaction	4	The numbers are reasonable, aligns with popular creative trends.
Loss Rate during Gameplay	38%	Loss rate during gameplay is higher than after gameplay, indicating that loss mainly happens during the game, possibly due to factors like high comprehension costs, high difficulty, and extended waiting times.
Loss Rate after Gameplay	22%	Loss rate during gameplay is higher than after gameplay, indicating that loss mainly happens during the game, possibly due to factors like high comprehension costs, high difficulty, and extended waiting times.
Loss Rate before Interaction	11%	Loss rate after interaction is higher than before, indicating that users are willing to interact but lose motivation to continue after that, leading to loss.
Loss Rate after Interaction	49%	Loss rate after interaction is higher than before, indicating that users are willing to interact but lose motivation to continue after that, leading to loss.

2）Scene-based Tracking

Tracking	Data	Data Interpretation（Orange = Key Conclusion）
Scene 1 - First Interaction Rate	60%	First interaction rate in Scene 1 is lower than Scene 2, indicating that users are more interested in Scene 2 with a higher willingness to interact. Therefore, the overall low first interaction rate can be primarily attributed to Scene 1.
Scene 1 - Rate of Redirection to Another Scene	60%	Since Scene 1 consists of only 1 step, the redirection rate to other scenes is the same as the first interaction rate.
Scene 1 - Redirect Rate	/	No redirection mechanism in Scene 1, no analysis.
Scene 1 - Loss Rate before Interaction	20%	No clear issues, and the loss rate before and after interaction is similar.
Scene 1 - Loss Rate after Interaction	18%	No clear issues, and the loss rate before and after interaction is similar.
Scene 2 - First Interaction Rate	95%	Excellent performance, with a significant improvement compared to Scene 1, indicating that this scene is attractive, and users enter it are likely to engage.
Scene 2 - Rate of Redirection to Another Scene	/	Scene 2 is the final scene and lacks a mechanism to redirection to other scenes. Therefore, no analysis is conducted.
Scene 2 - Redirect Rate	40%	No clear issues, still room for improvement.
Scene 2 - Loss Rate before Interaction	19%	The loss rate after interaction in Scene 2 is higher than before interaction. Therefore, the overall high loss rate after interaction can be primarily attributed to the lack of appeal during the Scene 2 gameplay process.
Scene 2 - Loss Rate after Interaction	32%	The loss rate after interaction in Scene 2 is higher than before interaction. Therefore, the overall high loss rate after interaction can be primarily attributed to the lack of appeal during the Scene 2 gameplay process.

✨Conclusions based on the data:

Issue 1：Scene 1 lacks appeal, leading to lower user interaction intent.

Optimization 1：Skip initial character selection and start directly from the core phase (Dressing), or adjust the creativity and visual effects of Scene 1 to enhance its appeal.

Issue 2：Long duration leads to many users lost midway. And the loss is concentrated after interaction in Scene 2.

Optimization 2：Reduce animation waiting times through acceleration or simplification to maintain the continuity of user actions. Additionally, enhance feedback on actions to improve the dressing experience in Scene 2.

II. Purpose B：A/B Test

💡If want to assess the A/B test results for multiple versions, you can directly focus on the stages related to the variables and compare the custom tracking data for that stage.

STEP 1: Compare the scene tracking of different versions to determine the best combination.

1）First, select the scenes and key metrics to focus on based on the goals of the A/B test.

For example, if the test goal is "Does adding guidance at the beginning make users more willing to interact", then it's essential to focus on the first interaction rate and the rate of redirection to other scenes in Scene 1 (definitions as the table above).

2）Using the interpretation approach above, select the version with the best data for each key metric.

3）Integrate the features associated with each optimal version to derive initial optimization directions.

4）*If custom tracking is set, refine your investigation using Step 2.

STEP 2: Verify optimization hypotheses by confirming relevant custom tracking data.

💡Custom tracking is defined by the content creator and closely linked to gameplay details, using the format "actionXX". It pinpoints gameplay milestones more accurately than basic tracking.

Approach to Viewing Custom Tracking:

Determine the meaning of each 'action,' including some linear and non-linear trackings.

• Linear：It refers to trackings with fixed timing. When arranged sequentially, you can calculate the loss rate and arrival rate for each custom stage.

• Non-Linear：It refers to independent buttons not tied to a fixed process. They can be compared with related linear trackings to calculate metrics such as failure rate and specific item usage rate.

Case 2：A/B Test

1）Click to preview the case A

Tracking	Date
👍Redirect Rate	61%
Redirect Rate during Gameplay	0%
Redirect Rate after Gameplay	61%
👍First Interaction Rate	88%
Game End Rate	63%
Replay Rate	/
TTE	5s
Duration	10s
Effective Interaction	1
Loss Rate during Gameplay	20%
Loss Rate after Gameplay	27%
Loss Rate before Interaction	30%
Loss Rate after Interaction	17%

2）Click to preview the case B

Tracking	Date
Redirect Rate	55%
Redirect Rate during Gameplay	0%
Redirect Rate after Gameplay	55%
First Interaction Rate	69%
Game End Rate	64%
Replay Rate	/
👍TTE	2.5s
Duration	12s
Effective Interaction	2
Loss Rate during Gameplay	21%
Loss Rate after Gameplay	28%
👍Loss Rate before Interaction	21%
Loss Rate after Interaction	26%

3）A/B Test Conclusion

Background：Creatives primarily test the effectiveness of different initial guidance methods, while other variables remain constant:

• Material A skips the initial movement step, plays an animation automatically, and guides the operation in the second step using an prop .

• Material B lacks text guidance during the prop usage stage.

Conclusion：

• Material A has a higher redirect rate and overall performs better.

• Material A exhibits a higher first interaction rate, showing that starting with a simple animation, skipping one interaction step, starting directly from entering the track yield the better result.

• However, in Material A, users show a lower TTE and a higher loss rate before interaction, which could be due to a higher comprehension cost of guidance and prolonged animation, leading to user disinterest.

• Both creatives have only one scene, so there's no need to split scenes for analysis.

✨Summing up the conclusions, the overall optimization direction is:

1）Add a animation at the beginning to immerse users in the ambiance. Keep the animation brief, within a 2-second duration, by accelerating or editing to prevent users from mistaking it for a video.

2）A single-step interaction is more effective than a two-step one. Consider transitioning to the store directly after the second interaction.

3）Provide clear and comprehensible instructions to avoid user delays in interaction due to a lack of understanding, thus reducing loss rate.