Menu

Data Fluctuation Troubleshooting

1. Revenue Fluctuation

(1) Calculation formula of income

Total revenue = per capita revenue * DAU

Per capita revenue = number of impressions per capita *eCPM/1000

 

(2) Three possible factors for changes in revenue:

① Changes in the number of impressions per capita;

② Changes in eCPM;

③ Changes in DAU

 

2. Fluctuation of impressions per capita

(1) Calculation formula for the number of impressions per capita

Number of impressions per capita = Number of traffic requests per capita * Traffic fill rate * Impression rate

Number of impressions per capita = number of ad requests per capita * ad fill rate * impression rate

(2) Recommended reference reports: TopOn Backstage - Funnel Report, TopOn Backstage - Full Report

(3) Use the Funnel report to check the date of change in per capita display:

Filter dimensions such as apps, placements, dates, and app versions, observe the dates when per capita impressions change, and locate the reasons based on factors such as product update changes and ad placement changes.

(4) Use funnel reports to investigate the reasons for changes in per capita impressions:

① Filter dimensions such as apps, placements, dates, etc., observe changes in users reaching each core link, and make corresponding adjustments

② Refer to the funnel report description, evaluate each link of the advertising impression, and adjust the product logic. Changes in various factors may cause fluctuations in per capita impressions.

(5) Troubleshooting process: refer to the calculation formula in (1) to investigate the causes of related factors.

The formula for calculating the number of impressions per capita:
Number of impressions per capita = Number of traffic requests per capita * Traffic fill rate * Impression rate
Number of impressions per capita = number of ad requests per capita * Ad fill rate * Impression rate
 

3. Fluctuation of per capita traffic requests

The number of traffic requests initiated by the application to TopOn SDK is one load

Troubleshooting ideas Explanation
Whether the app has been adjusted Whether the frequency of ad display has been adjusted; whether scenarios have been increased or reduced; whether ads are not displayed to paying users, etc.
Whether the delivery has been adjusted Is it due to fluctuations in user quality and user willingness to view ads?
TopOn background settings: whether to open or close the automatic request for ads, and retry if the request fails If automatic ad request is turned on, or if the request fails and is retried, the number of traffic requests will increase.
If you turn off the above options, it will be reduced.
It is recommended that developers who preload ads do not turn on the automatic ad request option.

 

4. Traffic fill rate fluctuations

Traffic fill rate: the fill rate of the traffic request initiated by the application itself to the TopOn SDK, that is, a load success.

Troubleshooting ideas Explanation
Check waterfall setting ● Whether to set the bottom layer--if not set, it is recommended to set the bottom layer or no reserve price layer. ● Whether the fill rate of each ad source, especially the bottom layer, fluctuates abnormally. If it is abnormal, it is recommended to check whether there is a problem with the ad source. Methods: ①Networks: Check the status of apps and ad sources. ②TopOn: Advanced-Log Analysis, and analyzing the error reasons and failure ratios of different ad sources. Find the cause of the error accordingly. ● Solution suggestions: ① Handle the status of networks apps and code bits according to the abnormal conditions. ②According to TopOn log analysis, the cause of the error report of the ad source should be solved accordingly. ③The traffic fill rate is low or becomes low: if the opening "request fails try again", add low-price or bottom ad sources.

 

5. Impression rate fluctuations

Troubleshooting ideas Explanation
Check related strategies that affect impression, such as whether to modify or adjust the impression strategy or timeout strategy.

● Whether the client timeout is set or adjusted, such as changing the original 5 seconds to 3 seconds. --It is necessary to confirm whether the client timeout policy and the timeout policy of TopOn background mediation-advanced settings are consistent.
● Whether the preload settings have been increased or decreased, etc.

● Whether the network is poor or the number of users disconnected from the network has increased, affecting the ad impression.
● Corresponding to the above possibilities, solutions:

① Adjust the client timeout, pay attention to the filling time of the placement and the ad source, as well as the client timeout. You can modify the TopOn background mediation-setting-waterfall flow filling waiting time to extend or shorten the waiting time.
② If the preloading strategy is added, the display rate is expected to be lower than when no preloading is set. This is because preloading will load in advance to ensure the user's ad display opportunities, and the loaded ads may not actually be displayed. Therefore, after adding preloading, the display rate becomes 40-60%, which is a reasonable data situation. Since it can increase user display opportunities, it has a positive impact on revenue.
③ If the number of users with weak or disconnected networks increases, as well as the number of users with low-end devices, ads may not be loaded, affecting the overall user display rate. It is recommended to adjust the user group for buying traffic. For such users who have too many applications, you can also add a strategy to preload ads in advance.

 

6. Fluctuations in ad requests per capita

Troubleshooting ideas Explanation
Changes in traffic requests If so, troubleshoot the cause of traffic request fluctuations - refer to the above steps
Whether the placements/ad sources have been adjusted/added or deleted Whether to open (create) or close (delete) the placement or ad source. If adjusted, it will affect the relevant data.

 

7. Ad fill rate Fluctuations

Networks filling rate: the filling status of advertising requests initiated by TopOnSDK to networks.

Troubleshooting ideas Explanation
Whether the reserve price of the ad source has been modified If you raise the reserve price, the fill rate will go down; if you lower the reserve price, the fill rate will go up.
Is the reason for network fluctuations Fluctuations in Network advertising resources, fluctuations in regional distribution, etc. For example, around the e-commerce festival, the budget of networks will increase. At the beginning of each year, eCPM will drop due to network budget updates.
Is it due to fluctuations in traffic quality Adjustment of delivery strategy causes fluctuations in user quality, etc.
Is it restricted by networks Are accounts, apps, and placement restricted/limited by networks

 

8. eCPM Fluctuations

Troubleshooting ideas Explanation
Whether to adjust the bottom price of networks Whether to increase or decrease the reserve price of networks
Whether the overall impression volume fluctuates abnormally If so, when the overall impression level increases or decreases sharply, eCPM will fluctuate greatly, which is a reasonable phenomenon.
Whether the impression of a certain layer fluctuates abnormally If so, eCPM is affected by the fluctuation of the impression of this layer, which is a normal phenomenon.
Whether to adjust the delivery strategy Fluctuations in user quality. When new users are added, the overall eCPM level will be higher.
Is it a special period? Special periods such as weekends and holidays cause price fluctuations
Whether the networks fluctuates Fluctuations in the advertising resources and budgets of networks are normal.
Is it restricted by networks Are accounts, apps, and placement restricted/limited by networks

 

9. DAU Fluctuations

There are fewer users, such as: fewer purchasing users, fewer organic users, fewer old users, etc.

Previous
Data gap troubleshooting
Next
Other Data Issues
Last modified: 2025-05-30Powered by