When someone buys from your store on an iPhone and chose "Ask App Not to Track," Meta records that a purchase happened. But it doesn't know who bought. It lost the email, the phone, the name, the city. The sale exists in your system and in Meta's report — but the data attached to it arrived incomplete. The problem isn't the sales Meta doesn't see. It's the data Meta loses on every sale it does see.
What is ATT and what changed with iOS 14?
In April 2021, Apple released iOS 14.5 with App Tracking Transparency (ATT), a privacy framework that forces every app on the App Store — including the Facebook app — to show a prompt asking the user whether they allow tracking across third-party apps and websites.
The screen is binary: "Allow" or "Ask App Not to Track." Most users choose not to allow it. According to Flurry Analytics studies, the global opt-in rate hovers around 21-25% — in other words, roughly 75-80% of iOS users block tracking.
When a user chooses not to allow tracking, their device's Identifier for Advertisers (IDFA) — the unique code Facebook uses to identify that user — comes back blank. Meta loses the ability to connect that click to a later purchase using the browser pixel.
To dig deeper into how the browser pixel works and why its limitations go beyond iOS, we recommend the article what the Meta pixel is and why it stopped working.
What data does Meta lose when a user says "no"?
Meta's browser pixel works like this: when someone clicks your ad and lands on your store, the pixel sets a cookie. If that person buys, the pixel reads the cookie, identifies the user and sends Meta the Purchase event data: email, name, phone, city, postal code, purchase value and the ID of the ad that drove the visit.
When iOS blocks tracking, that cookie can't be set properly or can't be read at the moment of purchase. The pixel can detect that a purchase happened — because it sits on your thank-you page — but it can't read the data of the user who bought. The sale reaches Meta with no email, no phone, no name, no city.
Match parameters are exactly that data: email, phone, first name, last name, city, state, postal code. Meta uses them to identify the user in its database. If that data doesn't arrive, Meta can't know whether the person who bought is the same one who saw your ad. To understand what these parameters are, we recommend the article on match parameters in Meta Ads.
The problem isn't that Meta doesn't count the sale. In Argentina, e-commerce stores usually see that the number of orders in Meta Ads and the number of orders in their store are fairly aligned. Meta records the sales. The real problem is that on many of those sales, Meta doesn't know who bought. It has the event, but not the profile. And without the profile, it can't build good lookalikes or properly optimize its delivery algorithms.
How does iOS affect your EMQ and your lookalikes?
Event Match Quality (EMQ) is the metric Meta uses to measure what percentage of the events you send it can link to an identified profile. A scale of 1 to 10: below 5 is low, above 7 is good.
When the pixel loses iOS users' match parameters, EMQ drops. Meta receives the Purchase event but without email or phone. That event adds to your EMQ's denominator but not to the numerator of matches. The score falls.
- Your lookalike audiences are built on incomplete profiles. The algorithm doesn't know exactly who bought, so the lookalike resembles who probably bought — with less certainty.
- Delivery optimization degrades. Meta delivers less to people who resemble your real buyers because it has fewer real buyers identified.
- Your CPA goes up because the algorithm needs more impressions to find the next buyer.
To dig deeper into how EMQ works, we recommend the article what EMQ is in Meta Ads and why it matters.
How much of your store's traffic comes from iOS?
In Argentina, iOS has a market share of between 20% and 35% depending on the product category. In fashion, accessories and cosmetics, the share of shoppers on iPhone can top 40%. That means a significant portion of your sales reach Meta with incomplete data.
You can check it in GA4: Reports → Tech → Overview → break down by operating system. And iOS isn't the only source of signal loss: ad blockers, Safari's privacy mode (ITP) and accelerated checkouts also contribute. iOS 14+ accelerated the problem and made it visible, but signal loss existed before.
The fix: send the data from your server, not the browser
Meta's Conversions API (CAPI) was designed to solve exactly this problem. Instead of relying on the browser pixel, CAPI sends purchase events from your server directly to Meta. The server always has the order data: email, phone, name, address.
With CAPI running, when someone buys from an iPhone with ATT blocked, the server still sends Meta the buyer's email and phone — hashed with SHA-256, privately. Meta does the matching on its server and connects that purchase to the user's profile. EMQ goes up because Meta can identify who bought, even though the browser pixel couldn't.
To understand how the Conversions API works in detail, we recommend the article Meta's Conversions API explained simply and the Escalafy server-side Pixel page.
How Escalafy solves the iOS problem — and goes one step further
Escalafy's server-side Pixel connects to your store (Tiendanube or Shopify) at the server level and sends every Purchase event to Meta with all available match parameters: the buyer's email, phone, name, city and postal code.
That data comes from the order data — not from the buyer's browser. iOS can't block it because it never passes through the user's device. The result is a higher EMQ and lookalike audiences built on complete data.
What sets Escalafy apart from other CAPI tools is that the event it sends Meta already includes the real profit of that sale — not just the revenue. The system deducts fees, product cost and shipping, and the event that reaches Meta carries the real net value. That lets the algorithm optimize toward the buyers who are most profitable for your business, not just the ones who spend the most. To see how it works, visit the Escalafy Pixel page.
Frequently asked questions about iOS, ATT and Meta Ads
Does iOS 14 also affect Android campaigns?
ATT is specific to iOS. Android has its own privacy system (Privacy Sandbox) that Google is rolling out gradually, but the impact on third-party tracking is smaller for now. The main impact of iOS is the loss of data from buyers who use iPhones — which in Argentina can be 20-40% of your customer base.
Does Meta's Aggregated Event Measurement solve the problem?
AEM is Meta's solution for reporting events under ATT with differential privacy. It lets you report up to 8 events per domain and uses statistical modeling to fill in the missing data. It helps, but it has limitations: modeled events are estimates, not real data. CAPI with server data is more accurate because it uses real order data.
Does implementing CAPI require technical knowledge?
A native CAPI implementation requires development work. Tools like the Escalafy Pixel offer the integration with no code: you connect your store, authorize access and the system does the rest in under 10 minutes.
How do I know if I'm already losing signal to iOS?
Go to Events Manager in Meta Ads and check the EMQ of your Purchase event. If it's below 6, you're losing signal. You can also check the diagnostics section in Events Manager for alerts about missing match parameters.
Can I ask my customers to turn tracking on?
You can't force anyone or actively ask for it — Apple doesn't allow it. The only technical alternative is to implement CAPI so order data reaches Meta from your server, without depending on device consent.
Data that arrives incomplete today costs you tomorrow.
The impact of iOS signal loss isn't immediate. It compounds. Every sale Meta records without an email or phone is a profile that doesn't make it into the lookalike, a buyer the algorithm doesn't learn to find, a CPA that rises a few cents every week. Six months later, the algorithm is optimizing with half the data it should have.
The technical fix exists and requires no code. What it does require is understanding that the problem isn't the sales Meta doesn't see — it's the data it loses on every sale it does see.
Keep reading
What Escalafy does for your store
Connect your channels and see the truth behind your numbers. No Excel, in real time.