Experiment
A Google ads experiment for users to experiment changes on multiple campaigns, compare the performance, and apply the effective changes.
Columns
| Name | Type | Behavior | Description |
| CustomerId | Long | ATTRIBUTE | Output only. The ID of the customer. |
| ExperimentDescription | String | ATTRIBUTE | The description of the experiment. It must have a minimum length of 1 and maximum length of 2048. |
| ExperimentEndDate | String | ATTRIBUTE | Date when the experiment ends. By default, the experiment ends on the campaign's end date. If this field is set, then the experiment ends at the end of the specified date in the customer's time zone. Format: YYYY-MM-DD Example: 2019-04-18 |
| ExperimentExperimentId | Long | ATTRIBUTE | Output only. The ID of the experiment. Read only. |
| ExperimentGoals | String | ATTRIBUTE | The goals of this experiment. |
| ExperimentLongRunningOperation | String | ATTRIBUTE | Output only. The resource name of the long-running operation that can be used to poll for completion of experiment schedule or promote. The most recent long running operation is returned. |
| ExperimentName | String | ATTRIBUTE | Required. The name of the experiment. It must have a minimum length of 1 and maximum length of 1024. It must be unique under a customer. |
| ExperimentOptimizeAssetsExperimentOptimizeAssetsExperimentSubtype | String | ATTRIBUTE | The subtype of the Optimize Assets experiment.
The allowed values are ADD_ASSETS_TO_ASSETLESS_RETAIL, ADD_VIDEO_ASSETS_TO_VIDEOLESS, COMPARE_ASSETS, UNKNOWN. |
| ExperimentPromoteStatus | String | ATTRIBUTE | Output only. The status of the experiment promotion process.
The allowed values are COMPLETED, COMPLETED_WITH_WARNING, FAILED, IN_PROGRESS, NOT_STARTED, UNKNOWN. |
| ExperimentResourceName | String | ATTRIBUTE | Immutable. The resource name of the experiment. Experiment resource names have the form: customers/{customer_id}/experiments/{experiment_id} |
| ExperimentStartDate | String | ATTRIBUTE | Date when the experiment starts. By default, the experiment starts now or on the campaign's start date, whichever is later. If this field is set, then the experiment starts at the beginning of the specified date in the customer's time zone. Format: YYYY-MM-DD Example: 2019-03-14 |
| ExperimentStatus | String | ATTRIBUTE | The Advertiser-chosen status of this experiment.
The allowed values are ENABLED, GRADUATED, HALTED, INITIATED, PROMOTED, REMOVED, SETUP, UNKNOWN. |
| ExperimentSuffix | String | ATTRIBUTE | For system managed experiments, the advertiser must provide a suffix during construction, in the setup stage before moving to initiated. The suffix will be appended to the in-design and experiment campaign names so that the name is base campaign name + suffix. |
| ExperimentSyncEnabled | Bool | ATTRIBUTE | Immutable. Set to true if changes to base campaigns should be synced to the trial campaigns. Any changes made directly to trial campaigns will be preserved. This field can only be set when the experiment is being created. |
| ExperimentType | String | ATTRIBUTE | Required. The product/feature that uses this experiment.
The allowed values are ADOPT_AI_MAX, ADOPT_BROAD_MATCH_KEYWORDS, AD_VARIATION, COMPARE_CAMPAIGNS, DISPLAY_AND_VIDEO_360, DISPLAY_AUTOMATED_BIDDING_STRATEGY, DISPLAY_CUSTOM, HOTEL_CUSTOM, OPTIMIZE_ASSETS, PMAX_REPLACEMENT_SHOPPING, PMAX_TEXT_CUSTOMIZATION_FINAL_URL_EXPANSION, SEARCH_AUTOMATED_BIDDING_STRATEGY, SEARCH_CUSTOM, SHOPPING_AUTOMATED_BIDDING_STRATEGY, SMART_MATCHING, UNKNOWN, YOUTUBE_CUSTOM. |
| ExperimentVideoExperimentVideoExperimentSubtype | String | ATTRIBUTE | The subtype of the Video experiment.
The allowed values are ASSET, ASSET_UPLIFT, DEMAND_GEN_ASSET, UNKNOWN. |
| Clicks | Long | METRIC | The number of clicks. |
| ClicksMarginOfError | Double | METRIC | The margin of error when estimating the experiment's effect on clicks. Together with clicks_point_estimate, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment / control - 1). This field specifies the radius of the confidence interval, which is centered at clicks_point_estimate. |
| ClicksPValue | Double | METRIC | The p-value for the null hypothesis that the experiment has no effect on clicks. Ranges from 0 to 1. Say if the p-value is 0.03, that means the probability of observing the data, if the experiment has no effect on clicks, is 3%. |
| ClicksPointEstimate | Double | METRIC | The point estimate when estimating the experiment's effect on clicks. Together with clicks_margin_of_error, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment / control - 1). This field specifies the point estimate, which is the center of the confidence interval: (clicks_point_estimate - clicks_margin_of_error, clicks_point_estimate + clicks_margin_of_error). |
| ControlClicks | Long | METRIC | The number of clicks on the control arm of an experiment. The treatment clicks value can be selected by using clicks. |
| ControlConversionValue | Double | METRIC | The conversion value metric on the control arm of the experiment. The treatment conversion value can be selected by using conversions_value. |
| ControlConversionValuePerCost | Double | METRIC | The conversion value per cost metric on the control arm of the experiment. The treatment conversion value per cost value can be selected by using conversions_value_per_cost. |
| ControlConversions | Double | METRIC | The conversions metric on the control arm of the experiment. The treatment conversions value can be selected by using conversions. |
| ControlCostMicros | Long | METRIC | The cost metric on the control arm of the experiment. The treatment cost value can be selected by using cost_micros. |
| ControlCostPerConversion | Double | METRIC | The cost per conversion metric on the control arm of the experiment. The treatment cost per conversion value can be selected by using cost_per_conversion. |
| ControlImpressions | Long | METRIC | The impressions metric on the control arm of the experiment. The treatment impressions value can be selected by using impressions. |
| ConversionValueChangePointEstimate | Double | METRIC | The point estimate when estimating the experiment's effect on conversion value change. Together with conversion_value_margin_of_error, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment / control - 1). This field specifies the point estimate, which is the center of the confidence interval: (conversion_value_change_point_estimate - conversion_value_margin_of_error, conversion_value_change_point_estimate + conversion_value_margin_of_error). |
| ConversionValueMarginOfError | Double | METRIC | The margin of error when estimating the experiment's effect on conversion value. Together with conversion_value_change_point_estimate, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment / control - 1). This field specifies the radius of the confidence interval, which is centered at conversion_value_change_point_estimate. |
| ConversionValuePValue | Double | METRIC | The p-value for the null hypothesis that the experiment has no effect on conversion value. Ranges from 0 to 1. Say if the p-value is 0.03, that means the probability of observing the data, if the experiment has no effect on conversion value, is 3%. |
| ConversionValuePerCostChangePointEstimate | Double | METRIC | The point estimate when estimating the experiment's effect on conversion value per cost change. Together with conversion_value_per_cost_margin_of_error, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment / control - 1). This field specifies the point estimate, which is the center of the confidence interval: (conversion_value_per_cost_change_point_estimate - conversion_value_per_cost_margin_of_error, conversion_value_per_cost_change_point_estimate + conversion_value_per_cost_margin_of_error). |
| ConversionValuePerCostMarginOfError | Double | METRIC | The margin of error when estimating the experiment's effect on conversion value per cost. Together with conversion_value_per_cost_change_point_estimate, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment / control - 1). This field specifies the radius of the confidence interval, which is centered at conversion_value_per_cost_change_point_estimate. |
| ConversionValuePerCostPValue | Double | METRIC | The p-value for the null hypothesis that the experiment has no effect on conversion value per cost. Ranges from 0 to 1. Say if the p-value is 0.03, that means the probability of observing the data, if the experiment has no effect on conversion value per cost, is 3%. |
| Conversions | Double | METRIC | The number of conversions. This only includes conversion actions which include_in_conversions_metric attribute is set to true. If you use conversion-based bidding, your bid strategies will optimize for these conversions. |
| ConversionsAbsoluteChangeMarginOfError | Double | METRIC | The margin of error when estimating the experiment's effect on conversions absolute change. Together with conversions_absolute_change_point_estimate, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment - control). This field specifies the radius of the confidence interval, which is centered at conversions_absolute_change_point_estimate. |
| ConversionsAbsoluteChangePValue | Double | METRIC | The p-value for the null hypothesis that the experiment has no effect on conversions absolute change. Ranges from 0 to 1. Say if the p-value is 0.03, that means the probability of observing the data, if the experiment has no effect on conversions absolute change, is 3%. |
| ConversionsAbsoluteChangePointEstimate | Double | METRIC | The point estimate when estimating the experiment's effect on conversions absolute change. Together with conversions_absolute_change_margin_of_error, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment - control). This field specifies the point estimate, which is the center of the confidence interval: (conversions_absolute_change_point_estimate - conversions_absolute_change_margin_of_error, conversions_absolute_change_point_estimate + conversions_absolute_change_margin_of_error). |
| ConversionsValue | Double | METRIC | The value of conversions. This only includes conversion actions which include_in_conversions_metric attribute is set to true. If you use conversion-based bidding, your bid strategies will optimize for these conversions. |
| ConversionsValuePerCost | Double | METRIC | The value of conversions divided by the cost of ad interactions. This only includes conversion actions which include_in_conversions_metric attribute is set to true. If you use conversion-based bidding, your bid strategies will optimize for these conversions. |
| CostMicros | Long | METRIC | The sum of your cost-per-click (CPC) and cost-per-thousand impressions (CPM) costs during this period. |
| CostMicrosChangePointEstimate | Double | METRIC | The point estimate when estimating the experiment's effect on cost change. Together with cost_micros_margin_of_error, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment / control - 1). This field specifies the point estimate, which is the center of the confidence interval: (cost_micros_change_point_estimate - cost_micros_margin_of_error, cost_micros_change_point_estimate + cost_micros_margin_of_error). |
| CostMicrosMarginOfError | Double | METRIC | The margin of error when estimating the experiment's effect on cost. Together with cost_micros_change_point_estimate, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment / control - 1). This field specifies the radius of the confidence interval, which is centered at cost_micros_change_point_estimate. |
| CostMicrosPValue | Double | METRIC | The p-value for the null hypothesis that the experiment has no effect on cost. Ranges from 0 to 1. Say if the p-value is 0.03, that means the probability of observing the data, if the experiment has no effect on cost, is 3%. |
| CostPerConversion | Double | METRIC | The cost of ad interactions divided by conversions. This only includes conversion actions which include_in_conversions_metric attribute is set to true. If you use conversion-based bidding, your bid strategies will optimize for these conversions. |
| CostPerConversionChangePointEstimate | Double | METRIC | The point estimate when estimating the experiment's effect on cost per conversion change. Together with cost_per_conversion_margin_of_error, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment / control - 1). This field specifies the point estimate, which is the center of the confidence interval: (cost_per_conversion_change_point_estimate - cost_per_conversion_margin_of_error, cost_per_conversion_change_point_estimate + cost_per_conversion_margin_of_error). |
| CostPerConversionMarginOfError | Double | METRIC | The margin of error when estimating the experiment's effect on cost per conversion. Together with cost_per_conversion_change_point_estimate, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment / control - 1). This field specifies the radius of the confidence interval, which is centered at cost_per_conversion_change_point_estimate. |
| CostPerConversionPValue | Double | METRIC | The p-value for the null hypothesis that the experiment has no effect on cost per conversion. Ranges from 0 to 1. Say if the p-value is 0.03, that means the probability of observing the data, if the experiment has no effect on cost per conversion, is 3%. |
| Impressions | Long | METRIC | Count of how often your ad has appeared on a search results page or website on the Google Network. |
| ImpressionsMarginOfError | Double | METRIC | The margin of error when estimating the experiment's effect on impressions. Together with impressions_point_estimate, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment / control - 1). This field specifies the radius of the confidence interval, which is centered at impressions_point_estimate. |
| ImpressionsPValue | Double | METRIC | The p-value for the null hypothesis that the experiment has no effect on impressions. Ranges from 0 to 1. Say if the p-value is 0.03, that means the probability of observing the data, if the experiment has no effect on impressions, is 3%. |
| ImpressionsPointEstimate | Double | METRIC | The point estimate when estimating the experiment's effect on impressions. Together with impressions_margin_of_error, they describe a confidence interval with a prescribed confidence level for the difference being estimated. The quantity being estimated is (treatment / control - 1). This field specifies the point estimate, which is the center of the confidence interval: (impressions_point_estimate - impressions_margin_of_error, impressions_point_estimate + impressions_margin_of_error). |
| Date | Date | SEGMENT | Date to which metrics apply. yyyy-MM-dd format, for example, 2018-04-17. |
| Period | String | SEGMENT | Predefined date range.
The allowed values are TODAY, YESTERDAY, LAST_7_DAYS, LAST_BUSINESS_WEEK, THIS_MONTH, LAST_MONTH, LAST_14_DAYS, LAST_30_DAYS, THIS_WEEK_SUN_TODAY, THIS_WEEK_MON_TODAY, LAST_WEEK_SUN_SAT, LAST_WEEK_MON_SUN. |
Pseudocolumns
Pseudocolumn fields are used in the WHERE clause of SELECT statements and offer more granular control over the data returned from the data source.
| Name | Type | Description |
| ManagerId | Long | Id of the manager account on behalf of which you are requesting customer data. |