How to Structure for Scale on Facebook - Part 1: The Ad Auction and Delivery System Explained

September 19, 2022

It’s no secret that the social advertising ecosystem (and the rest of the world) looks much different than it did just a few years ago. We know that many advertisers are searching for direction as they learn to navigate the constantly evolving landscape. 

Explaining the Ad Auction and Delivery System

Understanding how the Facebook ad auction works is the first step to adaptation. The ad auction formula consists of components that advertisers can leverage to increase their ad performance in the auction. The formula itself is a linear relationship between all components, making it easier to understand and predict how each variable will affect outcomes.

The Ad Auction

When advertisers create ads, they tell us who they want to show their ads to by defining a target audience. A person can fall into multiple target audiences. For example, one advertiser targets women who like skiing, while another advertiser targets all skiers who live in California. The same person (in this case, a female skier who lives in California) could fall into the target audience of both advertisers. When there's an opportunity to show someone an ad, the ads with a target audience that the person belongs to are eligible to compete in the auction.

To ensure that the winning ad maximizes value for both people and businesses, the winner of the auction is the ad with the highest total value. The total value is a combination of 3 major factors connected by the formula below:

  • Bid: The bid placed by an advertiser for that ad (in other words, what the advertiser is willing to pay to achieve their desired outcome).
  • Estimated action rates: An estimate of whether a particular person engages with or converts from a particular ad (in other words, the probability that showing an ad to a person leads to that desired outcome of the advertiser).
  • Ad quality: A measure of the quality of an ad as determined from many sources including feedback from people viewing or hiding the ad and and assessments of low-quality attributes in the ad, such as withholding information, sensationalized language and engagement bait.

Together, estimated action rates and ad quality measure ad relevance. Because these are components of the auction, an ad that’s more relevant to a person could win an auction against ads with higher bids.

The Ad Delivery System

Because the ad auction and ad delivery system behave like a supply-and-demand system, there are seasonal fluctuations and ecosystem changes that will affect the outcomes and costs of ad performance. There are two main causes of CPM fluctuations you may have experienced before:

  • Seasonal and Economic events:  The impact of seasonal and economic events on advertiser cost can be represented by supply and demand. Supply and demand typically have an inverse relationship, but multiple simultaneous forces on supply and demand can create different outcomes. Considering the impacts of seasonality and economic events in terms of supply and demand can help advertisers predict how this will translate into the ad auction. The examples provided by Meta below illustrate the expected impact of different macroeconomic variables and seasonality on the ad delivery system.

A positive economic event illustrates the typical inverse relationship between supply and demand. When supply (usership and user behavior) remains constant but we see advertisers competing for user attention by raising budgets, this leads to a flat level of impressions and thereby a higher cost per impression (impacting CPMs).

A relatable example of seasonality for advertisers is the holiday season. During this time, we see both supply and demand move as users are more actively searching for gifts to purchase, while advertisers are also spending more to capture this increased intent. In this example, purchases are likely to increase while CPAs remain constant.

A negative economic event in this example moves both the supply and demand curves. If users are spending more time on Facebook, the supply rises. But concerned advertisers also typically pull back spend during a negative economic event, pushing down demand. This resulted in increased impressions at a lower cost per impression, and subsequently lower CPM.

  • Effects on reporting due to user adoption of iOS tracking updates: With iOS 14.5+ adoption ramping up since July 2021, there has been an impact on advertising delivery and reporting, and performance fluctuations. While these effects have been widely felt, it’s important to recognize that these known reporting gaps do not mean our platform is not driving real business value for advertisers, or that your campaigns are not resulting in a purchase or major event. These changes just make it harder for us to measure the outcome. After extensive internal analysis, we’re able to publicly share that we now estimate - in aggregate - that we are currently underreporting iOS web conversions by approximately 8%, which is down from the 15% estimate we reported last September.

In part two of our collaboration, we’ll apply these concepts to strategies advertisers can use to structure for scale. Visit Meta for Business, for more information on the ad auction and delivery system. 

Elizabeth Kennedy

Marin Software
By submitting this form, I am agreeing to Marin’s privacy policy.

See why brands have relied on Marin to manage over $48 billion in spend