Ad Lift Test Results
Warning
Deprecation Notice
The Marketing Version 202401 (Marketing January 2024) has been sunset. We recommend that you migrate to the versioned APIs as well as migrate to the new Content and Community Management APIs to avoid disruptions. See the Migration page for more details.
If you haven’t yet migrated and have questions, submit a request on the LinkedIn Developer Support Portal.
Note
Please review the Getting Started Guide first to understand Lift Test concepts.
This is the resource for advertisers to view the result of a Ad Lift Test.
It can take several days for the surveys to begin delivering, so you won't immediately see results upon launching an Ad Lift Test.
Results can continue to fluctuate for up to 15 days after the end date as we continue to survey responses. This is caused by surveys that were served while the lift test was still running, but not viewed until after.
AdLiftTestResult Schema
| Field Name | Type | Description | Read only, create only |
|---|---|---|---|
| adLiftResultMetrics | AdLiftResultMetrics | Overall lift result metrics for this AdLiftKeyMetric (as opposed to breakdowns, like lift metric by survey response). This should contain all non-statistical metrics applying to the lift test. This will not be present if we don't have enough responses to calculate statistical significance. | Read only |
| adLiftResultStatisticalMetrics | AdLiftResultStatisticalMetrics | Statistical metrics about lift calculations for this AdLiftKeyMetric, like p-value and confidence intervals. This will not be present if we don't have enough responses to calculate statistical significance. | Read only |
| adLiftSurveyResponseBreakdowns | AdLiftSurveyResponseBreakdown[] | Raw response metrics for all survey answers (like YES or SOMEWHAT_FAVORABLE) for this key metric for both the control and test surveys. | Read only |
| adLiftTestResultKey | AdLiftTestResultKey | A composite key that uniquely identifies this brand lift result. | Read only |
| adLiftTestResultMetricMemberDemographicBreakdowns | AdLiftTestResultMetricMemberDemographicBreakdown[] | Ad Lift result metrics (ie relative lift, cost per lifted member, etc) broken down by professional demographic (ie members working for Linkedin or members working for companies in the logistics industry). For member privacy reasons, results will only be included for professional demographics with more than 10 responses. | Read only |
| adLiftTestSurveyResponseMemberDemographicBreakdowns | AdLiftTestSurveyResponseMemberDemographicBreakdown[] | Raw survey results (ie treatment clicks, control clicks) broken down by professional demographic (ie members working for Linkedin or members working for companies in the logistics industry). For member privacy reasons, results will only be included for professional demographics with more than 10 responses. | Read only |
AdLiftResultMetrics Schema
| Field Name | Type | Description | Read only, create only |
|---|---|---|---|
| absoluteLift | double | The difference in desired responses between the test and control groups, measured as an absolute difference in percentage points. For example, if 20% of control responses were desirable and 30% of test responses were desirable, then the absolute lift is 0.1. But if the numbers were reversed, then the value would be -0.1. This will always be in the range [-1.0, 1.0]. | Read only |
| costPerLiftedMember | MoneyAmount | The approximate cost of each additional person who we estimate would give a desirable response to your survey. Negative cost doesn't make sense in this context (we're not giving advertiser money back), so costPerLiftedMember will not be present if it would be negative. | Read only |
| relativeLift | double | The ratio of desired responses from the test group and desired responses from the control group, measured as a relative percentage. The range for this metric is unbounded. Relative lift can be over 1.0. For example, if desired responses tripled between control and test, it would be 3.0. It can also be negative if the test surveys received a lower percentage of desired responses than the control. | Read only |
| totalEstimatedLift | long | The estimated brand lift across the entire population that your campaigns reached. This value uses your total reach to extrapolate your survey results and estimate the improvement in the total number of people who would give a desirable response to your survey. | Read only |
AdLiftResultStatisticalMetrics Schema
| Field Name | Type | Description | Read only, create only |
|---|---|---|---|
| absoluteLiftConfidenceInterval | ConfidenceInterval | The confidence interval for absolute lift. | Read only |
| adLiftTestResultStrength | AdLiftTestResultStrength | A high-level measure of the strength of results generated based on p-value and power. This is aimed at advertisers who don't want to examine the detailed data. | Read only |
| pValue | double | The probability that this lift was caused by random chance, assuming that the null hypothesis is correct. (In this case, the null hypothesis is that there is no lift.) The p-value is in the range of [0, 1]. | Read only |
| power | double | The probability that the test rejects the null hypothesis when the alternative hypothesis is true. This is in the range [0, 1], where a lower number is a lower probability. | Read only |
| relativeLiftStatisticalStrengthMetrics | StatisticalStrengthMetrics | The statistical strength metrics related to the relative lift. | Read only |
StatisticalStrengthMetrics Schema
| Field Name | Type | Description | Read only, create only |
|---|---|---|---|
| ConfidenceInterval | ConfidenceInterval | The confidence interval for this metric. | Read only |
| adLiftTestResultStrength | AdLiftTestResultStrength | A high-level measure of the strength of results generated based on p-value and power. This is aimed at advertisers who don't want to examine the detailed data. | Read only |
| pValue | double | The probability that this lift was caused by random chance, assuming that the null hypothesis is correct. (In this case, the null hypothesis is that there is no lift.) The p-value is in the range of [0, 1]. | Read only |
| power | double | The probability that the test rejects the null hypothesis when the alternative hypothesis is true. This is in the range [0, 1], where a lower number is a lower probability. | Read only |
AdLiftSurveyResponseBreakdown Schema
| Field Name | Type | Description | Read only, create only |
|---|---|---|---|
| controlResponses | long | The number of responses for this answer from the control survey (sent to the control group, who have not seen an ad from the account under test). | Read only |
| surveyAnswer | AdLiftSurveyAnswer | The survey answer (e.g., YES, SOMEWHAT_FAVORABLE, etc.) that these metrics correspond to. | Read only |
| treatmentResponses | long | The number of responses for this answer from the treatment survey (sent to the treatment group, who have seen an ad from the account under test). | Read only |
AdLiftTestResultMetricMemberDemographicBreakdown
| Field Name | Type | Description | Read only, create only |
|---|---|---|---|
| adLiftResultMetrics | AdLiftResultMetrics | Ad Lift result metrics (ie relative lift, cost per lifted member, etc) for the given professional demographic. | Read only |
| adLiftResultStatisticalMetrics | AdLiftResultStatisticalMetrics | Statistical metrics (e.g. p-Value, confidence interval) about the lift test results for the given professional demographic. | Read only |
| adLiftTestDemographic | AdLiftTestMemberDemographic | The professional demographic these results are for (eg a job title of Software Engineer or a member company of 1001-5000 employers.) | Read only |
Note: Professional Demographic breakdowns are for directional insights and have not been tested for statistical significance.
AdLiftTestSurveyResponseMemberProfessionalDemographicBreakdown
| Field Name | Type | Description | Read only, create only |
|---|---|---|---|
| adLiftSurveyResponseBreakdown | AdLiftSurveyResponseBreakdown | Raw response metrics for a single survey answer (i.e. YES, SOMEWHAT_FAVORABLE, etc) for both the control and test surveys. | Read only |
| adLiftTestDemographic | AdLiftTestMemberDemographic | The professional demographic these results are for (eg a job title of Software Engineer or a member company of 1001-5000 employers.) | Read only |
Note: Professional Demographic breakdowns are for directional insights and have not been tested for statistical significance.
AdLiftTestMemberProfessionalDemographic
| Field Name | Type | Description | Read only, create only |
|---|---|---|---|
| demographicCategory | AdLiftTestMemberDemographicCategory | A professional demographic category like MEMBER_COMPANY_SIZE | Read only |
| demographicName | MultiLocaleString | Translations for the professional demographic value. For example a MEMBER_JOB_SENIORITY professional demographic might be "Entry" and a MEMBER_COMPANY_SIZE professional demographic might be "1001-5000 employees". Translations will be provided for all locales supported by LMS. This field is derived from the raw professional demographic value (which isn't exposed directly by this resource). | Read only |
AdLiftTestMemberProfessionalDemographicCategory Enum
| Name | Description | Example Value |
|---|---|---|
| MEMBER_JOB_FUNCTION | Job function | Human Resources |
| MEMBER_JOB_SENIORITY | Job seniority | Director |
| MEMBER_COMPANY_INDUSTRY | Company industry | Oil & Energy |
| MEMBER_COMPANY_SIZE | Company size | 1001-5000 employees |
| MEMBER_AGE_RANGE | Age range | 18-24 |
| MEMBER_GENDER | Gender | Female |
AdLiftTestResultKey Schema
| Field Name | Type | Description | Read only, create only |
|---|---|---|---|
| adLiftKeyMetric | AdLiftKeyMetric | The ad lift key metric (example: BRAND_RECALL) that this result is associated with. | Read only |
| adLiftTestUrn | AdLiftTestUrn | The ad lift test that this result is associated with. | Read only |
ConfidenceInterval Schema
| Field Name | Type | Description | Example Value | Required | Read only, create only |
|---|---|---|---|---|---|
| lowerBound | double | The inclusive lower bound of the confidence interval. | -0.02 | Required | Read only |
| upperBound | double | The inclusive upper bound of the confidence interval. | 0.05 | Required | Read only |
FindByAdLiftTests Finder
This finder can be used to request the results for an Ad Lift Test. It returns a list of AdLiftTestResults.
The request must include a header for Restli Protocol 2.0.0 as X-Restli-Protocol-Version: 2.0.0
Restli 2.0 requires URNs in query params to be URL encoded. For example, urn:li:adLiftTest:1245678 would become urn%3Ali%3AadLiftTest%3A1245678
Query Parameters
| Parameter | Type | Description | Required |
|---|---|---|---|
| adLiftTests | Array of Ad Lift Test Urns | The ad lift tests to fetch results for | Yes |
Sample Request
GET https://api.linkedin.com/rest/adLiftTestResults?q=adLiftTests&adLiftTests=List(urn%3Ali%3AadLiftTest%3A{liftTestId})
GET https://api.linkedin.com/v2/adLiftTestResults?q=adLiftTests&adLiftTests=List(urn%3Ali%3AadLiftTest%3A{liftTestId})
Sample Response
AdLiftTestResults will return an array containing all relevant results. If an Ad Lift Test is requested for which the member doesn't have access, it will be omitted from the results. If the request contains no valid Ad Lift Test Results, an empty array will be returned.
{
"elements": [
{
"adLiftSurveyResponseBreakdowns": [
{
"controlResponses": 50,
"treatmentResponses": 100,
"surveyAnswer": "YES"
},
{
"controlResponses": 50,
"treatmentResponses": 100,
"surveyAnswer": "NO"
},
{
"controlResponses": 50,
"treatmentResponses": 100,
"surveyAnswer": "NOT_SURE"
}
],
"adLiftTestResultKey": {
"adLiftKeyMetric": "AWARENESS",
"adLiftTestUrn": "urn:li:adLiftTest:12345"
},
"adLiftResultMetrics": {
"totalEstimatedLift": 150000,
"absoluteLift": 0.09234,
"relativeLift": 0.21233
},
"adLiftResultStatisticalMetrics": {
"pValue": 0.012342,
"absoluteLiftConfidenceInterval": {
"upperBound": 0.1231,
"lowerBound": -0.1233
},
"power": 0.4132131,
"relativeLiftStatisticalStrengthMetrics": {
"adLiftTestResultStrength": "MEDIUM",
"ConfidenceInterval": {
"lowerBound": 0.1231,
"upperBound": -0.1233
},
"pValue": 0.012342,
"power": 0.4132131
},
"adLiftTestResultStrength": "MEDIUM"
}
}
],
"paging": {
"count": 10,
"start": 0,
"links": []
}
}
A successful request will return a 200 OK response code.