Viewing performance data in CSJ reports

CSJ reports show how your ads perform in terms of impressions, clicks, CTR, CPM, and revenues on the app-, ad placement- and account-level, respectively.

（1）Estimated Revenue (RMB/USD)

The estimated revenue of your account during a selected period.

This is an estimated value, which might be updated before your revenue is confirmed at the end of each month.

（2）Impressions

The number of times ads are effectively shown.

This does not include repeated impressions, timed-out impressions, or impressions generated by "test" status placements.

The definitions of effective impressions for different ad formats are as follows:

Splash Ads: ad display height should be no less than 75% of the screen height; ad display width should be the same as the screen width; and the display time should be more than 1s (Splash Ads unavailable for global traffic);

Native Ads: take up more than 20% of the total pixels; the display time should be more than 1s;

Interstitial Video Ads & Rewarded Video Ads: video starts playing;

Repeated impressions: when the ad returned in response to a successful request is shown to the user multiple times, the repeated impression(s) will be ignored and only count as 1 Impression;

Timed-out impressions: when the ad returned in response to a successful request is shown to the user after specific time limit, it will not be counted towards the number of Impressions (time limit: 3 hours for Splash Ads, 1 hour for other ad formats);

Impressions generated by "test" status placements: when a placement is set to "test" mode, the ad returned in response to a successful request from this placement will not be counted towards the number of Impressions;

（3）Clicks

The number of effective clicks.

This does not include repeated clicks or clicks generated by "test" status placements.

Detailed description:

Repeated clicks: when the ad returned in response to a successful request is clicked by the user multiple times, the repeated click(s) will be ignored and they are counted as 1 Click;

Clicks generated by placement of "test" status: clicks on placements set to "test" mode will not be counted as effective clicks by CSJ;

When a placement is set to "test" mode, the clicks generated from ad in response to a successful request from this placement will not be counted towards the number of Clicks;

（4）CTR

CTR= (Clicks/Impressions)*100

（5）eCPM

eCPM = Estimated Revenue*(1000/Impressions)

（6）Requests

The number of requests received by CSJ.

A request is counted as 1 even if it asked for more than one ad creatives.

The pre-cached requests automatically initiated by CSJ are counted towards the number of Requests (only for Splash Ads and Rewarded Video Ads).

This does not include the error requests reported by SDK. When a request initiated by user device encountered a -12 error, the request will not be reported to CSJ server and thus will not be counted towards the number of Requests;

Pre-cached requests automatically initiated by CSJ:

Splash Ads: when an user device initiates a request, CSJ will initiate a pre-cached request simultaneously and the number of Requests will be counted as 2;

Rewarded Video Ads/Interstitial Video Ads/Interstitial Ads: when a video finishes playing and a callback is made, CSJ will automatically initiate a pre-cached request, which will be counted towards the number of Requests;

Examples of pre-cached video requests:

User device initiates a request for Rewarded Video; the current number of Requests is 1.

When the video ad creative returned by CSJ finishes playing and makes a callback, CSJ will initiate a pre-cached request; the current number of Requests is 2.

If the video ad creative returned in response to this pre-cached request is successfully loaded and haven t expired by the time of next Rewarded Video Ad, the pre-cached video can be used and the end device will not initiate another request; the current number of Requests remains 2.

If the pre-cached video fails to be loaded or expires, no video creatives will be available for the next Rewarded Video Ad and the end device will initiate another request; the current number of Requests will then become 3.

（7）Responses

The number of ad responses returned by CSJ.

A response is counted as 1 even if CSJ returned more than one ad creatives.

Responses for the pre-cached requests automatically initiated by CSJ are counted towards the number of Responses.

（8）Fill Rate

Fill Rate = (Responses/Requests)*100%

（9）Ad Requests

The number of requested ad creatives received by CSJ.

This measures the requested number of ad counts included in ad requests. Different ad formats have different ad count limits. For example, a Native Ad can request up to 3 ad counts, and a Banner Ad can only request 1 ad count.

When a request is received for X ad counts, it will be counted as X Ad Requests.

The number of ad counts included in the pre-cached requests automatically initiated by CSJ are counted towards the number of Ad Requests.

For example:

When a user device initiates a request for Splash Ad, CSJ will initiate a pre-cached request simultaneously (with identical ad count of 1); the current number of Ad Requests will be 2.

（10）Ad Responses

The number of ad creatives returned by CSJ .

The Ad Responses are counted as X when a request asks for Y ads creatives and CSJ successfully returns X ad creatives.

The number of ad creatives returned in response to the pre-cached requests automatically initiated by CSJ are counted towards the number of Ad Responses.

（11）Ad Fill Rate

Ad Fill Rate = (Ad Responses/Ad Requests)*100%

（12）Ad Impression Rate

Ad Impression Rate = (Impressions/Ad Responses)*100%

Why does my eCPM fluctuate?

Fluctuations in eCPM are determined by CTR, CPC, traffic quality, and multiple other factors. In general, when the traffic volume is stable, CTR should also remain stable.

Possible causes of eCPM fluctuations include:

- Changes in advertising competition that lead to fluctuations in CPC, thus affecting eCPM;

- Advertisers budgets;

- Fluctuations in the magnitude of large-volume traffic flows.