A Comprehensive Guide to Meta Conversions API (C-API): Effective Ad Optimization with Server-Side Tracking

With the privacy changes post iOS 14.5, it's no longer sufficient to send data (events) to ad platforms like Meta using only a browser pixel as increased privacy limits the amount of user data that can be tracked through browser-based methods. This will inevitably lead to a less accurate identification of your target audiences and sub-optimal ad auctions, making your ad targeting less effective. Which is why Server Side Tracking is a must now and as an Advertiser, you must also send the same events server-side using the conversion API (C-API) of the ad platform (e.g. Meta, Tiktok and Google Ads etc).

While Meta's C-API is a standard framework that accepts relevant events, the way you surface, organize, and send your valuable 1st party and Zero-party (customer shared) data through the C-API significantly influences its effectiveness. Hence, various C-API integrations have been designed to help ad platforms identify your target audience and assist the ad auction in determining who should see your ad, when, and how often.

But choosing the right C-API integration can be confusing and overwhelming, which is why in this article we will discuss different C-API solutions, and help you choose one that is ideal for your business.

Let’s dive in!

Steps for Surfacing Events

Before we discuss different C-API integrations, let’s talk about different steps involved in surfacing server-side tracked events to the ad platforms for auction optimization and audience targeting.


Step 1 - Identify

Collect events from websites, apps, and backend ordering systems, and identify known users who are not signed in during the current session and thus appear as anonymous users.


Step 2 - Predict

As an advertiser, you run paid media campaigns to achieve certain business objectives such as -acquiring new customers, acquiring leads that eventually convert through CRM nurturing, and acquiring high-value customers who will repurchase within the next 3-6 months, etc.

The value of all events you capture varies. For example, not every user who clicks on your ad and visits your site is likely to convert, not just in the current session, but also not convert in the next 7 or 28 days.

Similarly, not all leads you generate will convert through CRM. But, if you can identify high-value leads that are more likely to convert and predict this as soon as you capture that lead, then such information is highly valuable for ad auction optimization.

So, if your goal is to acquire customers who are likely to repurchase in the next 3-6 months (belonging to higher-LTV groups) and you are willing to pay a bit more to acquire such customers so you can improve your LTV:CAC, you will need a reliable prediction of LTV (pLTV) as soon as someone becomes your customer for the first time, or even better, if you have such prediction available even before someone buys from you.


Step 3 - Enrich

From step 1, identity resolution is used for session stitching - connecting marketing identifiers and user identifiers to the events. This makes the event payload rich with as detailed user information as possible. This allows ad platforms to map the event to their user graph and use the event for ad auction optimization and attribution (reporting).


Types of C-API implementations

Now that we understand the steps involved in surfacing events with C-API, let’s talk about different C-API solutions.


Basic C-API

Out of the box E-commerce platform integrations like Shopify C-API, BigCommerce C-API etc., as well as Customer Data Platforms (CDP) integrations (e.g. Segment) belong to this category. They merely pass through events as is, with very limited identity and enrichment capabilities. Because of this, you typically see limited event payload for upper and mid funnel events such as add to cart or page view. E-commerce C-API implementations are generally free for advertisers to use and comes with an easy setup process through an intuitive UI.


Advanced C-API

Most of the 3rd party C-API implementations fall in this category. These solutions perform identity resolution using a graph built off advertiser’s 1st-party data to re-identify past customers and visitors, and enrich event payloads with stitched user and marketing identifiers.


Predictive C-API

This is a state-of-the-art implementation of C-API. It differentiates itself from the advanced C-API implementation in the following ways:

  • The Identify step uses a 3rd party consumer graph, where available (geographically), in addition to the advertiser’s 1st party graph. This allows for confidence-graded matching (i.e., identifying past customers with a higher degree of confidence and reduces the chance of mis-identification). Also, this allows us to fill in additional user identifiers, for example, a user may use one email address for e-commerce purchases and a different email for Facebook login. With a 3rd party consumer graph, we can send additional user emails and get a match which otherwise would be missed.

  • Prediction is the other differentiation, where instead of sending all events, you could send high-value events contextual to the marketing objective. You need the power of predictions

    • to identify movable middle prospects with a conversion model,

    • predictive LTV for value optimization, and

    • predictive lead scores for running lead campaigns to acquire leads that ultimately convert


How do you select the right C-API solution for your business?

We hope the above explanations have given you a clearer understanding of different C-API solutions.

When selecting the right C-API for your business, consider the following points:

  • A mere improvement in Event Match Quality (EMQ) is not enough, as improving EMQ is necessary but not a complete solution to improving ad performance.

  • Conduct A/B tests before making a final selection from different C-APIs available. Where possible, setup a new data source (pixel) so that you can isolate testing of different C-APIs.

  • Ensure you have the configuration capabilities to orchestrate events that align with your business objectives and campaign setup. For instance, if your goal is to acquire new customers, align platform optimization with last touch attribution, or conduct Meta conversion lift studies, your C-

    API solution should be configured accordingly.


Not all C-API created equal

Below, we have aggregated various C-API solutions, categorized into the three C-APIs types discussed above, so you can choose the one that's ideal for your business.


What some brands have achieved with Angler's Predictive C-API

Now that you know what different C-API solution are capable of, allow us to share what some of our partner brands have been able to achieve with Angler AI’s Predictive C-API.


Case Study: The impact of shifting from Shopify native C-API to Angler’s predictive C-API

An apparel brand that wanted to acquire new customers using Meta paid ads. They transitioned from Meta's native C-API to Angler's predictive C-API, optimizing their daily spend with Angler events. Despite similar daily expenditures before and after the switch, this brand saw a 26% improvement in Marketing Efficiency Ratio (MER) for new customers after they made the switch. This improvement reached a higher, steady state almost immediately.


Case Study: Improving 7-day ROAS with Angler’s predictive C-API identifying the moveable middle

Another consumer brand selling women's clothing switched from their in-house C-API, maintained by their engineering team, to Angler's predictive C-API. Angler was using a 7-day conversion model to identify users who are likely to convert between day 2 and day 7 after engaging with an ad. This brand reported comparable performance on 1-day click purchases, however, a 58% higher ROAS with Angler's predictive C-API. This improvement came from the movable middle segment of their market that converted between day 2 and day 7. This trend continued beyond day 7 as well, with an 81% higher ROAS on 28-day click attribution with a 58% lower cost per order (CPO).


Case Study: A consumer brand improved leads to conversion by 55% with flat cost per lead using Angler’s predictive C-API

A consumer brand offering home and backyard design services transitioned from their Customer Data Platform (CDP), Segment C-API to Angler's predictive C-API. This brand sells services costing $1,500 or more, with a long consideration cycle. For them, a majority of the conversions occur between day 8 and 28, which is outside of their ad attribution window. The brand was running Meta ads optimized towards lead captures, where users sign up for CRM after completing a design survey. Angler's predictive model identified the top 20% of leads that have a higher than average conversion rate.

And, when the brand switched their ad optimization to high-value leads (custom event optimization) using Angler’s Predictive C-API, they improved lead conversion by 55% at the same cost per lead with Meta Custom Event Optimization, and Reels-ready creatives. This was demonstrated via concurrent campaigns for control and test cells. Month-over-month, their cost per lead (CPL) improved by 24% (pre vs post-test); and year-over-year, CPLs dropped by more than half (-53%) while ROAS increased by 92% (Q1'23 vs Q1'24).


To Conclude

Server-side tracking with Conversions API (C-API) has become a crucial aspect for effective ad optimization given the current advertising landscape with increased privacy regulations. The 3 types of C-API solutions discussed in this article, including Basic, Advanced, and Predictive, offer different levels of consumer data enrichment and identification capabilities. Although the right solution will depend on your business needs and objectives, we highly recommend conducting A/B tests to ensure you have the right configuration capabilities to orchestrate events that align with your business objectives.

At the end of the day, no matter which C-API you choose, it's clear - you absolutely need a C-API solution and server-side tracking. As we continue to navigate the post-iOS 14.5 world, it's becoming more essential to send the best data to ad platforms. This way, your ads can find better customers who love your brand, making your ads more profitable and increasing revenue for your brand.