You advertise on Facebook or Instagram and notice that something isn't right with the numbers. You feel something is missing, but you don't know exactly what. Or perhaps you've already noticed one of the following suspicious phenomena:
- Your campaigns are getting more expensive, but you don't understand why – your ads haven't gotten worse, yet your CPA is increasing.
- You've heard of the "Conversions API", but you don't know exactly what it is, or if you need to deal with it.
- Meta shows significantly fewer conversionsthan your CRM or e-commerce admin interface.
- You have many iOS users, and you suspect their conversions are simply disappearing from your measurements.
- Your visitors are running AdBlockers, and you're concerned this is distorting your data.
- Your remarketing and lookalike campaigns are performing worse and worse, and you don't understand why your audiences are dwindling.
If even one of these sounds familiar, this article is for you.
We'll show you what's behind these problems, what Meta CAPI is, how it differs from Pixel, and what can – and should – be done to remedy the situation.
.png)
Introduction - Why Are Your Numbers Lying?
Imagine the following scenario: you set up a Facebook campaign, optimize your ads, and monitor your metrics. Meta's ad manager shows that the campaign generated 40 purchases last week. Then you open your webshop's admin panel or CRM, and you see 67 orders from the same source.
The discrepancy isn't due to incorrect measurement, but incomplete tracking, as nearly 40% of your conversions are invisible to Meta.
As a marketing agency, we can confirm that this is far from a rare or isolated incident, and it's not Meta's fault in the way many people perceive it. This is a systemic issue that has gradually emerged over the past few years and now affects every Facebook advertiser to some degree.
So, what's causing this invisibility problem?
The answer is short and surprisingly simple: the world of web analytics has changed. Previously, digital measurement relied almost exclusively on browser-based solutions, primarily cookies, which was a simple yet effective system. A user would click on an online ad, the browser would store the cookie, and then upon conversion, the cookie would report it to the advertising platform. This mechanism worked well and was sufficient for years.
However, the digital world has changed significantly since then: in 2021, Apple released its iOS 14.5 update, introducing the App Tracking Transparency (ATT) framework. From that point on, every iOS app had to request explicit user consent for data tracking. The vast majority of iPhone users – an estimated 60-85% – declined this, instantly making their activity invisible to advertisers.
In parallel, the number of ad blocker users has grown: today, 30-40% of users run some form of ad-blocking extension in their browser, which blocks not only ads but also tracking scripts – including the Facebook Pixel. Furthermore, the built-in privacy features of Safari and Firefox browsers (ITP - Intelligent Tracking Prevention) further restrict the lifespan and effectiveness of cookies.
#m1-y#The GDPR and other data protection regulations are leading to an increasing number of users rejecting cookie acceptance.##
And what's the result of all this? Traditional Pixel-based tracking is getting more and more holes.
The lost data is invisible; it simply disappears. Consequently, Meta's algorithm tries to optimize using incomplete data, which leads to higher advertising costs, weaker targeting, and distorted ROAS.
But what's the solution then? How can we access our lost measurement data?
#m1-p#The solution is none other than the Meta Conversions API, commonly known as CAPI.##
What benefits does CAPI bring in practice?
Implementing CAPI doesn't just mean theoretical improvements or cleaner data. Its impact is truly seen in tangible, measurable business results.
[SEGSEGMENT 9]
is truly evident.
20-40% more measured conversions
After implementing CAPI, the most frequently observed and cited result is that clients typically see 20-40% more conversions in the ad manager than before – without having changed anything in their campaigns or any alteration in previous sales figures.
This does not mean that there was suddenly a 20-40% increase in purchases; it merely means that conversions previously invisible to the Pixel are now accessible to Meta.
So, reality hasn't changed; only our measurement has become more accurate.
Lower CPA and CPL
When Meta's algorithm can work with more and more accurate data, it can optimize more effectively.
This directly reduces advertising costs, meaning that with the same budget, more conversions are achieved, resulting in lower CPA and CPL (customer acquisition cost).
The operation is simple: previously, the algorithm tried to determine which users our ads should be shown to based on incomplete data.
The
more accurate picture provided by CAPI, however, gives the algorithm better/greater insight into who actually converts, allowing it to target these users more precisely. Better Audiences
Meta's algorithm no longer just uses simple audiences; it continuously learns from conversion data.
Since CAPI provides the algorithm with more and more accurate conversion data, this directly improves its learning process.
As a result, finding similar users becomes more efficient, and CAPI helps include users in the targeting scope who could not be tracked by browser-side measurement – thus making remarketing audiences more complete and accurate.
Future-proof measurement
The trend towards increasingly strict data protection regulations and a cookie-less online presence is unstoppable.
Third-party cookies are gradually disappearing from most browsers, users are becoming more privacy-aware, and the regulatory environment is also becoming stricter.
#m1-y#In this context, CAPI is not just a useful addition, but also an essential cornerstone of future-proof measurement.##
It allows you to build an infrastructure that doesn't depend on cookies, iOS permissions, or browser settings. If you implement CAPI now, future browser and data privacy restrictions won't catch you off guard. Meanwhile, you'll gain a data measurement advantage that late-reacting competitors will find difficult to catch up with.
What is the Facebook Pixel, and why is it no longer enough on its own?
Before we dive into the world of CAPI, it's worth understanding exactly what the Facebook Pixel is, how it works, and why it has encountered its current limitations.
%201.png)
How the Pixel works
The Facebook Pixel is a short JavaScript code snippet, which you embed into your website's HTML code. When a visitor arrives on your page, the Pixel loads in the browser and starts monitoring events:
- which page they visited,
- what they added to their cart,
- what they purchased,
- which contact form they filled out, etc.
The Pixel sends these events directly from the user's browser to Meta's servers – the process therefore takes place exclusively on the user's device and through their browser. occurs.
#m1-y#This is known as client-side data sending.##
For identification, the Pixel largely relies on cookies:
When someone first visits the site (and accepts cookies), the pixel saves a unique identifier (cookie) in their browser. When the visitor returns later – even days later – the Pixel recognizes them by this identifier, links it to their visits, and finally records their conversion for the appropriate advertising campaign.
So, when is the Pixel still effective?
The Pixel works excellently in cases where the user:
- has accepted cookies,
- is not running an ad blocker,
- is not using an iOS device,
- and the browser in question does not restrict third-party scripts.
In the above cases, the Pixel accurately records events, reliably identifies the user, and transmits data to Meta in real-time.
#m1-y#Furthermore, another advantage of the Pixel is that it is relatively simple to implement: you just need to insert the code once and it works.##
It does not require deep technical knowledge or continuous developer intervention.
Latent Data Loss
The biggest problem with Pixel-based measurement is that data loss is invisible. You don't get an error message, and no red light flashes in either the ad manager or the event manager. The Pixel silently loses some of the data, while you only see what it saw: the remaining, fragmented picture.
You don't know how much is missing, and that's precisely what's dangerous. Since Meta's algorithm also doesn't know it's optimizing from incomplete data, it makes increasingly expensive and less effective decisions.
What is Meta CAPI (Conversions API)?
The Meta Conversions API - CAPI for short - is a server-side data sending solution developed by Facebook.
Its name accurately describes its essence:
#m1-p#An application programming interface (API), through which you - the advertiser - can send conversion event data directly from your own server to Meta's system.##
The essence of server-side data sending
Unlike the Pixel, CAPI does not communicate through the user's browser with Meta, but rather your web server (or backend system) connects directly to Meta's API and sends events to it.
%202.png)
You can think of it this way: the Pixel only gives you as much insight into your store's real traffic as if you were observing it from the storefront. You can't see into every aisle, only the parts of the store visible from the window. In contrast, CAPI is like your store's internal camera system, allowing you to monitor even the most hidden corners.
What data can CAPI transmit?
Through CAPI, you can essentially send in the same events/event data, as with the Pixel: page view, add to cart, purchase, registration, lead submission, initiate checkout, etc.
#m1-p#Furthermore, with CAPI, you can send events that the Pixel cannot measure at all. ##
Examples of such events include offline conversions or even CRM events, such as lead qualification or contract signing.
Why are CAPI's data flows not affected by firewalls and cookie blockers?
Since CAPI's data flow is entirely server-to-server based , the user's browser plays no role in it.
The data flow is therefore as follows:
- An event occurs on the website
- → Your server records this event
- → The server initiates an API call to Meta.
It is clear that this process involves no browser scripts, no cookies, and no client-side components that could be blocked or restricted by an ad blocker, cookie consent, or iOS.
%203.png)
Firewalls do not pose an obstacle in this case either, as server-to-server communication typically occurs via HTTPS, which corporate network filters generally allow.
The Measure of Effectiveness
Now that we understand how CAPI works, it's worth discussing an important metric that becomes visible with the introduction of CAPI.
#m1-p#This metric is none other than Event Match Quality (EMQ), also known as Event Matching Quality Score.##
This indicator is a value measured on a 0-10 scale, showing how effectively/reliably Meta can match individual events to users' Facebook/Instagram profiles.
The more identifiable parameters you send to Meta (e.g., email address, phone number, browser ID, IP address, user ID), the higher your EMQ will be, and the more accurately Meta can attribute the given conversion to the appropriate ad interaction. Ideally, an EMQ value above 6 is good, and above 8 signifies excellent match quality.
It is important to emphasize that identifiable data (email, phone number, IP, etc.) is not stored raw by Meta, but is transmitted to Meta already hashed (encrypted), thus in a privacy-compliant manner it is transmitted.
How to Integrate CAPI into Your Site?
There are several possible ways to implement CAPI, and the appropriate solution always depends on the specific website's platform, technical infrastructure, and requirements. We distinguish three main approaches:
- Platform Plugin or Native Integration
Many popular e-commerce platforms (Shopify, WooCommerce, Unas, Shoprenter, etc.) have with built-in Meta CAPI integration, which can be implemented relatively quickly, without significant developer work. This approach for simpler webshops typically provides a sufficient starting point.
- Google Tag Manager server-side tagging (sGTM)
This is a more flexible, intermediate solution that leverages the GTM infrastructure to implement server-side data sending. It is particularly useful if you want to send server-side data to multiple advertising platforms (Meta, Google, TikTok) simultaneously from a single system.
For this to work, a partner or infrastructure capable of server-side data processing (e.g., Stape.io or Google Cloud) is also needed to ensure data reception and forwarding.
- The most in-depth solution is direct developer (API) integration
In this case, Meta's API is called directly from the backend system, which provides the greatest flexibility and best data quality.
This approach is justified when a custom-developed system is running in the background, or offline and CRM events also need to be measured.
#promobox-newsletter-en###
Pixel vs. CAPI
Now that we understand the operating principles of both technologies, let's look at their main similarities and differences.
%204.png)
Why not one OR the other, but both?
The most important takeaway for today:
#m1-y#Pixel and CAPI are not substitutes for each other, but rather complements.##
The best thing you can do is run both simultaneously: the Pixel records what it can, and CAPI fills in the gaps for cases the Pixel cannot see. Together, they provide a much more complete picture than either could offer alone. CAPI is not enough on its own, because although it reliably sends server-side data, it doesn't see everything that happens in the browser (e.g., clicks, interactions), so some important events might be missed. The Pixel sees precisely these browser-side events, so together they provide a complete, reliable picture.
Deduplication - avoid double counting
By now, you might be wondering, okay, but won't I be recording most of the data twice?
To prevent this, there is a mechanism called deduplication, during which a unique identifier (event_id) is assigned to each event, which is transmitted by both the Pixel and CAPI, and Meta recognizes events with the same event_id, thereby counting them only once.
It is clear from the above that correct deduplication setup is crucial, because without it, data becomes distorted and campaign optimization is compromised.
Summary and final thoughts
If you've read our article this far, then you already have a good understanding of what Meta CAPI is about and why it's important.
But how do you know if your measurements are accurate?
Most advertisers don't know for sure. They just feel that campaigns are getting more expensive, the numbers don't add up, and Meta shows less than what they see in their CRM. Exactly what is being lost and why can only be revealed by a thorough measurement audit.
In our free CAPI audit, we will examine your current measurement configuration, identify weaknesses, and provide concrete recommendations on which CAPI solution would best suit your system, without obligation.
Every day spent procrastinating means lost data. If you are unsure, Contact us!
#promobox-en#Do you know if your measurements are accurate? If you are unsure, request our free audit service.##
