How does Meta use your CAPI signals?

Congratulations! You've successfully set up Meta's Conversion API (CAPI) and your server-side integration is live. 🎉 But what happens after your marketing event data is sent through Meta’s CAPI? Let’s break down the next steps and how to get the most out of your integration.

Step 1: Identification

The first thing Meta does with your data is to match it against its vast user database. This process is key for determining who sees your ads. The accuracy of this match depends on the quality and completeness of your data — details like email addresses, phone numbers, and Facebook Click IDs are essential.

Meta grades the quality of your match on a scale of 1 to 10, known as the Event Match Quality (EMQ). A higher score indicates a better match between your data and Meta’s user profiles, increasing the likelihood of your ads reaching the right people.

Improving your EMQ score can feel like an obsession, and while it’s important, it's just one piece of the puzzle. Let’s move on to the next critical phase.

Step 2: Auction Training & Optimization

Once Meta matches your event data, the platform uses this information to participate in ad auctions. Meta uses a variant of Vickrey-Clarke-Groves (VCG) auction

Meta’s Estimated Action Rate is a critical component that helps you win auctions. Estimated Action Rate determines how likely users are to take your desired action, such as a purchase, signup, or download. For example, if you’re optimizing for purchases, Meta assesses how likely a user is to make a purchase within your attribution window. Your CAPI data plays a crucial role in this assessment and impacts your overall auction outcomes.

Here’s how it works:

• Meta uses your CAPI data to decide whether to bid on a user during an auction.
• The platform calculates how much to bid based on the likelihood of that user converting,
• You win the auction and your ad is shown to that user if the total value of your ad is highest amongst all advertisers participating in that auction.

The total value is a combination of 3 major factors (source) :

• 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). There are multiple ways to manage your bid in the ad auction. Learn more about bid strategies.
• 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). Note: Engaging in clickbait and engagement bait does not improve ad performance.
• 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 assessments of low-quality attributes in the ad, such as withholding information, sensationalized language and engagement bait.

Step 3: Platform Attribution

After the auction process, your CAPI data also helps Meta measure the performance of your campaigns, ads, and creatives. In this stage, Meta uses your data to evaluate how successful your campaigns have been in driving the desired outcomes.

While identification and attribution are important, Auction Training & Optimization is where the real impact happens. This is the process where Meta learns how to bid on the right audiences and improve the effectiveness of your campaigns.

How to Optimize Your Auction Training

To get the most out of your Meta campaigns, focus on how you train the auction system. Your goals will define your approach:

• Reduce Customer Acquisition Costs (CAC): Train the auction to target your brand’s “movable middle” — people who are likely to convert but haven’t yet.
• Increase ROAS or 90-Day LTV: Target users who are likely to spend more on their first purchase or repurchase within 90 days.
• Boost Free Trial Conversions: Focus on high-quality signups who are most likely to convert into paying customers.
• Lower Cost per Converted Lead: Optimize for leads that are likely to convert in the next 28-90 days.

A Pro Tip on EMQ Scores

While it’s tempting to chase higher EMQ scores, this shouldn’t be your only priority. Winning on EMQ but failing to train the auction effectively can lead to missed opportunities for growth. Similarly, obsessing over platform attribution (your campaign’s “final grade”) won’t help if you haven’t properly trained the auction first.

‍

How Angler AI Can Help

At Angler AI, we obsess over Auction Training & Optimization. Our predictive CAPI solution uses your zero-party and first-party data to train Meta (and other ad platforms) to deliver the results you need. Our approach ensures that your in-platform attribution aligns with your preferred model and stays trustworthy, so you can make confident, data-driven decisions.

Let’s win the auction training battle together!

By following these steps, you can maximize the impact of your CAPI integration and ensure that Meta’s ad platform works in your favor.

Ready to optimize your campaigns? Book a demo today, and let's get started!

‍