How Ad Quality is Evaluated in Meta Ads: Breakdown of the Auction System and Quality Ranking

When you run ads in Meta (Facebook/Instagram), it's not a simple purchase of impressions. Every time a user opens their feed, Meta conducts an instantaneous ad auction to decide which sponsored ad will appear. This happens billions of times a day across all of the company's platforms.

Unlike traditional auctions where the highest bidder always wins, Meta's system works differently. Your bid is only one of the factors. The quality of your creative can be more important than the money spent.

Meta's quality assessment system is not static—it is constantly evolving and improving through machine learning and the analysis of millions of user interactions daily. Every day, the system analyzes the behavior of billions of users, allowing it to more accurately predict and evaluate the quality of advertising materials.

How the Auction Principle Works

Every time there is an opportunity to show your ad to someone in your target audience on Meta platforms (Facebook, Instagram, Messenger), it enters an auction. The system gathers all ads that include this person in the advertiser's chosen audience and moves them to the auction stage.

The winning creative will be the one with the highest Total Value. It's important to understand that this isn't just the highest bid—the system considers a complex indicator that includes several key components.

Total Value Formula

According to official Meta information, Total Value consists of three main factors:

Bid — the amount an advertiser is willing to pay to show their ad. This can be an automatic bid set by Meta's algorithm or a manual bid defined by the advertiser.

Estimated Action Rate — an estimate of how likely a specific person is to complete the advertiser's desired action, such as visiting a website or installing an app. Meta uses machine learning to predict this based on user behavior both on and off Meta platforms.

Ad Quality — a comprehensive assessment of the ad's overall quality, which includes user feedback and assessments of low-quality advertising attributes.

Three Components of Ad Relevance Diagnostics

Meta replaced the previous Relevance Score with three more detailed diagnostic tools that allow media buyers to better understand their ads' performance. These three components are available in Ads Manager for ads that have received at least 500 impressions.

1. Quality Ranking

Quality Ranking explains how your ad's perceived quality compares to other ads competing for the same audience. Meta measures ad quality through various signals, including feedback from people viewing or hiding the ad, as well as assessments of low-quality attributes.

Low-quality attributes include:

• Too much text in the ad image
• Sensationalized language (clickbait)
• Engagement bait — direct calls to like, comment, or share
• Misleading or inaccurate content

2. Engagement Rate Ranking

Engagement Rate Ranking shows how your ad's expected engagement rate compares to competing ads targeting the same audience. This indicator predicts how likely people are to interact with your ad through likes, comments, shares, or clicks.

3. Conversion Rate Ranking

Conversion Rate Ranking explains how your ad's expected conversion rate compares to ads with the same optimization goal competing for the same audience. The expected conversion rate calculates the probability that a person who viewed your ad will complete your optimization goal.

How Machine Learning Works in Meta Ads

Machine learning is a system that learns when receiving new data, without being explicitly programmed, to perform complex tasks quickly and efficiently. Meta uses machine learning to generate predicted action rates and assess ad quality.

To determine the predicted action rate, machine learning models predict the likelihood that a specific person will perform the advertiser's desired action based on the business goal selected for the ad.

Data Analyzed by the System:

• On-Facebook Behavior: actions a person takes within Facebook apps (e.g., clicking an ad or liking a post).
• Off-Facebook Behavior: actions taken outside Facebook that businesses share via Business Tools (e.g., visiting a website, purchasing, or app installation).
• Ad Content: visual and textual elements of the advertisement.
• Time of Day: when the ad is shown.
• Interactions: how people overall interact with the ads.

Factors Influencing Ad Quality

1. Pre-auction Assessment

Before entering an auction, every ad undergoes an initial review. This includes automated content analysis and may include manual moderation.

• Advertising Standards: compliance with policies regarding prohibited categories.
• Technical Quality: file size, resolution, video quality.
• Textual Content: checks for clickbait, grammar, and misleading info.
• Visual Quality: text-to-image ratio, readability, design professionalism.

2. Factors During the Auction

• Audience Relevance: how well the ad matches a user's interests and patterns.
• Historical Performance: previous effectiveness of your ads, account, and page.
• Landing Page Quality: loading speed, mobile optimization, and relevant content.
• Contextual Factors: device type, location, and auction competition.

Conclusion

The ad quality assessment system in Meta Ads is a critical multi-factor mechanism continuously improved through machine learning. The main takeaway for advertisers: ad quality has as much impact on auction success as your bid. To understand and improve this quality, use the three diagnostic tools: Quality, Engagement, and Conversion Rate Ranking. Meta's goal is to provide a positive user experience by showing relevant and high-quality ads—when you meet this goal, you get lower costs, more reach, and better results.