News for Developers

Back to News

Announcing Ad Relevance Score

Christine Lu
February 11, 2015

Today we are launching Ad Relevance Score and also making it available via the Marketing API. Ads Relevance Score is a metric that provides an estimate of how relevant your ad is to its target audience. The more relevant an ad is to your audience, the better it’s likely to perform. To give you more insight into the performance of your ads, we are adding the relevance score as a metric into the ad report stats API.

In the report stats API, relevance_score has been added as a data_column, only applicable when queried with adgroup_id.

use FacebookAds\Object\AdAccount;

$account = new AdAccount('act_<AD_ACCOUNT_ID>');

$params = array(
'date_preset' => 'last_14_days',
'data_columns' => array(
'adgroup_id',
'relevance_score',
),
'sort_by' => 'relevance_score:score',
'sort_dir' => 'desc',
'filters' => array(
array(
'field' => 'relevance_score:score',
'type' => '>',
'value' => 2,
),
array(
'field' => 'relevance_score:negative_feedback',
'type' => 'starts_with',
'value' => 'LOW',
),
),
);

$stats = $account->getReportsStats(array(), $params);
from facebookads.objects import AdAccount
account = AdAccount('act_<AD_ACCOUNT_ID>')

params = {
'date_preset': 'last_14_days',
'data_columns': [
'adgroup_id',
'relevance_score',
],
'sort_by': 'relevance_score:score',
'sort_dir': 'desc',
'filters': [
[
'field': 'relevance_score:score',
'type': '>',
'value': 2,
],
[
'field': 'relevance_score:negative_feedback',
'type': 'starts_with',
'value': 'LOW',
],
],
}

stats = account.get_report_stats(params=params)
curl -G \
-d 'date_preset=last_14_days' \
-d 'data_columns=["adgroup_id","relevance_score"]' \
-d 'sort_by=relevance_score:score' \
-d 'sort_dir=desc' \
-d 'filters=[{field:"relevance_score:score",type:">",value:2},{field:"relevance_score:negative_feedback",type:"starts_with",value:"LOW"}]' \
-d 'access_token=<ACCESS_TOKEN>' \
https://graph.facebook.com/act_<AD_ACCOUNT_ID>/reportstats

Response

The relevance_score is a JSON dictionary of the following keys and values:

key type value

score

int

1-10, or empty if not enough impressions (< 500 impressions)

positive_feedback

string

A rating based on the number of times your audience took a desired action after seeing your ad, such as shared or liked it, or helped you achieve your objective, such as visiting your website. When positive feedback is high it means people are responding well to your ad.
One of "Low", "Medium", "High" or field is not returned if not enough impressions
"Low" means (1 <= score <= 2)
"Medium" means (3 <= score <= 6)
"High" means (7 <= score <= 10)

negative_feedback

string

A rating based on the number of times your audience hid your ad or otherwise indicated a negative experience such as choosing not to see ads from you. When negative feedback is high it means people don't want to see this ad.
One of "Low", "Medium", "High" or field is not returned if not enough impressions
"Low" means (1 <= score <= 2)
"Medium" means (3 <= score <= 6)
"High" means (7 <= score <= 10)

status

string

one of "OK", "NOT_ENOUGH_IMPRESSIONS", "UNKNOWN_ERROR"
NOT_ENOUGH_IMPRESSIONS – There were not enough impressions served for Facebook to determine a relevance score
UNKNOWN_ERROR - Try again later

Sample response:

{"data":
[{"adgroup_id": <ADGROUP_ID_1>,     
"date_start":"2015-01-13",
"date_stop":"2015-01-13",
"relevance_score":{
"score":10,
"positive_feedback":"HIGH",
"negative_feedback":"LOW",
"status":"OK"
}
},
{"adgroup_id":<ADGROUP_ID_2>,
"date_start":"2015-01-12",
"date_stop":"2015-01-12",
"relevance_score":{
"score":10,
"positive_feedback":"HIGH",
"negative_feedback":"LOW",
"status":"OK"
}
},
...
}

Here's an example of a response with not enough impressions:

"data": [   
{
"adgroup_id": "&amp;lt;ADGROUP_ID_1>",
"date_start": "2014-10-26",
"date_stop": "2014-10-26",
"relevance_score": {
"status": "NOT_ENOUGH_IMPRESSIONS"
}
},
...

relevance_score

cannot be queried on the ad set, campaign, or account level. Not compatible with placement, cross device columns, age, gender, or country breakdowns.

How Relevance Score is calculated

Ads that are served more than 500 times start to receive a daily Relevance Score between 1 and 10. The score is calculated based on the positive and negative feedback we expect an ad to receive from its target audience. The inputs vary depending on the ad's objective, but may include:

  • Positive Feedback: The more clicks, likes and shares we expect your ad to get, the higher its score will be.
  • Negative Feedback: The more times we expect people to hide or report your ad, the lower its score will be.

The score is updated as people interact and provide feedback on the ad.

The Relevance Score will not impact delivery for campaigns with a pre-defined delivery through Reach and Frequency.

How to use Relevance Score

It can lower the cost of reaching people.

Put simply, the higher an ad's relevance score is, the less it will cost to be delivered. This is because our ad delivery system is designed to show the right content to the right people, and a high relevance score is seen by the system as a positive signal.

Of course, relevance isn't the only factor our ad delivery system considers. Bid matters too. For instance, if two ads are aimed at the same audience, there's no guarantee that the ad with an excellent relevance score and low bid will beat the ad with a good relevance score and high bid. But, overall, having strong relevance scores will help advertisers see more efficient delivery through our system.

It can help advertisers test ad creative options before running a campaign.

Advertisers can test different combinations of images and copy with different audiences, and learn which combinations offer the highest relevance scores.

It can help optimize campaigns already in progress.

While ad campaigns are running, advertisers can monitor their relevance scores. If a score begins to dip, it may be an indicator that the ad's creative or audience needs to be refreshed.

How to improve Relevance Score

Read up on our best practices for improving your ad's relevance.


Related News

September 15 2016

December 07 2015

November 11 2015

October 26 2015

October 21 2015

October 07 2015