Good Questions, Real Answers: How Does Facebook Use Machine Learning to Deliver Ads?
Good Questions, Real Answers: How Does Facebook Use Machine Learning to Deliver Ads?
Facebook BusinessDelivering personalized ads maximizes value for both people and businesses. It helps businesses reach customers affordably, grow and create jobs, and provides people a better experience. Our latest Good Questions, Real Answers post will help you understand how Facebook delivers this value through ad auctions that use machine learning to determine which ads to show to people on our apps.
How does Facebook decide which ads to show people?
We determine which ads to show people based on two main factors: audience targeting selected by advertisers and the results of our ad auction.
First, advertisers choose their target audience through our self-service tools. Audiences are created based on categories like age and gender, as well as actions people take on our apps such as liking a Facebook Page or clicking on an ad. Advertisers can also use information they have about their audience, like a list of emails or people who’ve visited their website, to build a custom audience or a lookalike audience.
Next, when determining which ads to show someone, our system gathers ads that include that person in the advertiser’s chosen audience. These ads move to the auction stage.
For ads that enter the auction, Facebook selects the top ads to show to a person based on which ads have the highest total value score — a combination of advertiser value and ad quality. We find advertiser value by multiplying an ad’s bid by the estimated action rate. This is an estimate of how likely that particular person is to take the advertiser’s desired action, like visiting the advertiser’s website or installing their app. We then add the ad quality score, which is a determination of the overall quality of an ad. We use machine learning to inform this process, as we explain below.
What is machine learning and how does Facebook use it to inform ad delivery?
Machine learning is a system that learns as it receives new data, without being explicitly programmed, to carry out complex tasks quickly and efficiently. Facebook uses machine learning to generate the estimated action rate and the ad quality score used in the total value equation.
To find the estimated action rate, machine learning models predict a particular person’s likelihood of taking the advertiser’s desired action, based on the business objective the advertiser selects for their ad, like increasing visits to their website or driving purchases. To do this, our models consider that person's behavior on and off Facebook, as well as other factors, such as the content of the ad, the time of day, and interactions between people and ads.
- Examples of behaviors on Facebook that the models consider include things a person does while using Facebook apps, like clicking on an ad or liking a post.
- Examples of behaviors off Facebook that the models consider include things a person does outside of Facebook that businesses share with us via our Business Tools, like visiting a website, purchasing a product or installing an app.
To generate an ad’s quality score, our machine learning models consider the feedback of people viewing or hiding the ad, as well as assessments of low-quality attributes (like too much text in the ad's image, sensationalized language or engagement bait).
The advertiser’s bid, the estimated action rate and the ad quality score are combined to calculate the ad’s total value score in the ad auction.
How does machine learning improve ad delivery?
Over time, as more people view an ad, share feedback on it or click through to make a purchase on an advertiser’s website, our models get better at predicting the estimated action rate and ad quality. Since billions of people use our apps and engage with ads each day, our system gets lots of information to help improve its calculations, furthering our ultimate goal of maximizing value for both people and businesses.
Ads with the highest bid don’t always win the auction. Ads with lower bids often win if our system predicts a person is more likely to respond to them, or finds that they’re higher quality. This allows businesses of all sizes to compete in the auction and reach customers on any budget.
What controls are available to people to help determine what ads they see?
Ad Preferences
The Ad Preferences page is a place for users to review and update their ad settings so they can take more control over what information we use when deciding what ads to show them. People can opt out of seeing ads based on data from partners, and can also hide or report any ad from any advertiser with a few taps. Manage your ad preferences
Why Am I Seeing This?
Why Am I Seeing This shows users information about the detailed targeting options the advertiser chose to reach them. It lets them tap on posts and ads in New Feed, get context on why they’re appearing and take action to further personalize what they see.
Off-Facebook Activity
Some businesses send us information about users’ activity on their sites and we use that information to show them ads that are relevant to them. Off-Facebook Activity lets users see a summary of that information and clear it from their account if they want to.
What are common misunderstandings about Facebook Ads?
- Facebook does not sell people’s data to advertisers or anyone else.
- We don’t share information with advertisers that personally identifies individuals unless they’ve given us permission.
- Facebook does not use the content of people’s text messages or phones’ microphone to inform ads or to change what they see in News Feed.
At Facebook, our goal is to create personalized, data-driven ad experiences that are interesting and useful for people and effective for businesses. When we’re able to show someone the right ad for the right product, it’s valuable for everyone, and value is what good personalized advertising is all about.
Get Facebook Business news in your inbox.
Sign up for our monthly newsletter for the latest updates, insights, marketing trends and articles from Facebook.
Tags
Announcements Small Business Agency Large Business
Announcements
June 11, 2020
Good Questions, Real Answers: How Does Facebook Use Machine Learning to Deliver Ads?
D
How does Facebook decide which ads to show people?
We determine which ads to show people based on two main factors: audience targeting selected by advertisers and the results of our ad auction.
First, advertisers choose their target audience through our self-service tools. Audiences are created based on categories like age and gender, as well as actions people take on our apps such as liking a Facebook Page or clicking on an ad. Advertisers can also use information they have about their audience, like a list of emails or people who’ve visited their website, to build a custom audience or a lookalike audience.
Next, when determining which ads to show someone, our system gathers ads that include that person in the advertiser’s chosen audience. These ads move to the auction stage.
For ads that enter the auction, Facebook selects the top ads to show to a person based on which ads have the highest total value score — a combination of advertiser value and ad quality. We find advertiser value by multiplying an ad’s bid by the estimated action rate. This is an estimate of how likely that particular person is to take the advertiser’s desired action, like visiting the advertiser’s website or installing their app. We then add the ad quality score, which is a determination of the overall quality of an ad. We use machine learning to inform this process, as we explain below.
What is machine learning and how does Facebook use it to inform ad delivery?
Machine learning is a system that learns as it receives new data, without being explicitly programmed, to carry out complex tasks quickly and efficiently. Facebook uses machine learning to generate the estimated action rate and the ad quality score used in the total value equation.
To find the estimated action rate, machine learning models predict a particular person’s likelihood of taking the advertiser’s desired action, based on the business objective the advertiser selects for their ad, like increasing visits to their website or driving purchases. To do this, our models consider that person's behavior on and off Facebook, as well as other factors, such as the content of the ad, the time of day, and interactions between people and ads.
- Examples of behaviors on Facebook that the models consider include things a person does while using Facebook apps, like clicking on an ad or liking a post.
- Examples of behaviors off Facebook that the models consider include things a person does outside of Facebook that businesses share with us via our Business Tools, like visiting a website, purchasing a product or installing an app.
To generate an ad’s quality score, our machine learning models consider the feedback of people viewing or hiding the ad, as well as assessments of low-quality attributes (like too much text in the ad's image, sensationalized language or engagement bait).
The advertiser’s bid, the estimated action rate and the ad quality score are combined to calculate the ad’s total value score in the ad auction.
How does machine learning improve ad delivery?
Over time, as more people view an ad, share feedback on it or click through to make a purchase on an advertiser’s website, our models get better at predicting the estimated action rate and ad quality. Since billions of people use our apps and engage with ads each day, our system gets lots of information to help improve its calculations, furthering our ultimate goal of maximizing value for both people and businesses.
Ads with the highest bid don’t always win the auction. Ads with lower bids often win if our system predicts a person is more likely to respond to them, or finds that they’re higher quality. This allows businesses of all sizes to compete in the auction and reach customers on any budget.
What controls are available to people to help determine what ads they see?
Ad Preferences
The Ad Preferences page is a place for users to review and update their ad settings so they can take more control over what information we use when deciding what ads to show them. People can opt out of seeing ads based on data from partners, and can also hide or report any ad from any advertiser with a few taps. Manage your ad preferences
Why Am I Seeing This?
Why Am I Seeing This shows users information about the detailed targeting options the advertiser chose to reach them. It lets them tap on posts and ads in New Feed, get context on why they’re appearing and take action to further personalize what they see.
Off-Facebook Activity
Some businesses send us information about users’ activity on their sites and we use that information to show them ads that are relevant to them. Off-Facebook Activity lets users see a summary of that information and clear it from their account if they want to.
What are common misunderstandings about Facebook Ads?
- Facebook does not sell people’s data to advertisers or anyone else.
- We don’t share information with advertisers that personally identifies individuals unless they’ve given us permission.
- Facebook does not use the content of people’s text messages or phones’ microphone to inform ads or to change what they see in News Feed.
At Facebook, our goal is to create personalized, data-driven ad experiences that are interesting and useful for people and effective for businesses. When we’re able to show someone the right ad for the right product, it’s valuable for everyone, and value is what good personalized advertising is all about.
Related Articles
Small Business ·
How Does Facebook Use Machine Learning to Deliver Ads?
Delivering personalized ads helps businesses reach customers. Understand how Facebook delivers value through ad auctions that determine which ads people see.
Announcements ·
Allowing the Promotion of Non-Medical Masks on Facebook
In March, we temporarily banned ads and commerce listings for masks. Now, we're scaling back this ban to allow people to promote and trade non-medical masks.
Announcements ·
Celebrating Asian and Pacific Islander Heritage Month: A Livestream Panel Discussion With World-Renowned Leaders
This panel brings together leaders in the Asian community to discuss the unique challenges facing Asian entrepreneurs and the broader community through COVID-19.
Get Facebook Business news in your inbox.
Sign up for our monthly newsletter for the latest updates, insights, marketing trends and articles from Facebook.
