Safari blocks third-party cookies immediately and deletes first-party cookies after 7 days. If a buyer visits your store from Safari on a Wednesday and buys on Wednesday the following week, the browser pixel no longer has the cookie to identify them. Meta records the sale but doesn't know who it was. You didn't lose the sale — you lost the data on who bought, which is what Meta needs to build your audiences and optimize your campaigns.
What is Safari's ITP and why does it matter for your e-commerce store?
ITP stands for Intelligent Tracking Prevention, a technology Apple introduced in Safari in 2017 and has expanded with every version of the operating system. Apple's stated goal is to protect its users' privacy by limiting tracking tools' ability to follow behavior across sites. The side effect for advertisers is a progressive loss of attribution data.
It's not a minor problem. Safari is the dominant browser on iPhone and iPad, and in LATAM the iPhone has high market penetration in the upper-middle income segment — which is often exactly the ICP of many e-commerce stores. According to 2025 usage data, Safari accounts for between 25% and 35% of web traffic in markets like Argentina, Mexico and Colombia.
How does ITP affect the Meta pixel, step by step?
The Meta pixel is JavaScript code that runs in the visitor's browser and writes a cookie to identify them on future visits. ITP affects this in three layers:
| Cookie type | What ITP does | Impact on the pixel |
|---|---|---|
| Third-party cookies (cross-site) | Blocked since 2020 | The pixel can't track users across domains |
| First-party cookies (external script) | Expire after 7 days | Sessions that span more than 7 days are lost |
| localStorage as an alternative | Also capped at 7 days | The usual workaround doesn't work either |
The most common case is first-party cookies with a 7-day expiration. A user who visits your store from a Meta ad, considers the purchase for several days and finally buys between week 1 and week 2 — normal behavior in higher-priced categories — shows up in Meta as a conversion with no identification data.
How many of your buyers come through Safari?
You can see it in Google Analytics 4 or in your store's dashboard: look for the browser breakdown of sessions over the last 30 days. If between 25% and 45% of your traffic is Safari and your typical consideration window is longer than 3 days, ITP is eating a significant share of your conversion data.
And this is independent of iOS 14 / ATT. The confusion is common: ATT affects native apps (Instagram, Facebook) when the user taps "ask app not to track." ITP affects web browsers — Safari on iPhone, iPad and Mac — without the user having to do anything. They're two different mechanisms and both take data away from you.
Why can't the Meta pixel alone solve ITP?
The Meta pixel is code that runs in the user's browser. ITP controls exactly what that code can store and for how long. Meta knows this is a problem and has published guides to mitigate it, but none of them solve it from the browser — because the restriction is imposed by the browser.
The only solution that doesn't depend on the browser is server-side tracking: instead of the pixel trying to store a cookie that Safari will delete in 7 days, your server identifies the user with its own data (registered email, customer account) and sends the event to Meta through the Conversions API. The server has no ITP. Read more about how it works in Server-side tracking: what it is and why you need it if you run Meta Ads.
What data does Meta lose when ITP deletes the cookie?
When the tracking cookie no longer exists at the moment of purchase, the Purchase event sent by the browser pixel arrives without the matching parameters Meta needs to identify the buyer:
- The buyer's email — without it, Meta can't match the conversion to an Instagram or Facebook account.
- The phone number — the second most valuable signal for matching.
- Name, city and postal code — data that raises EMQ when combined with email or phone.
The result is a low EMQ. Meta knows a purchase happened but doesn't know who made it. It can't include that person in your lookalikes, can't exclude them from your prospecting campaigns and can't optimize bidding for purchase value. Lookalike audiences built on incomplete data are less precise, and CPA rises over time.
How does the Escalafy Pixel solve the ITP problem?
The Escalafy Pixel runs on your store's server, not in the buyer's browser. When a sale is confirmed, Escalafy reads the buyer's data from the Tiendanube or Shopify backend — email, phone, name, address — and sends it to Meta through the Conversions API. ITP can't delete what lives on the server.
Unlike other CAPI solutions that only send the purchase event, the Escalafy Pixel includes the real profit figure for each order — with product cost, platform fees, installments and shipping costs already deducted. Meta receives not just "someone bought" but "someone bought and the profit was X pesos." That lets you optimize for real profitability, not gross revenue. See how it works in Escalafy server-side pixel.
Frequently asked questions about Safari ITP and Meta Ads tracking
Does ITP also affect Chrome and Firefox?
Chrome and Firefox don't implement ITP — it's an Apple-specific mechanism. However, Firefox has blocked third-party cookies by default since 2019, and Chrome removed third-party cookies for a growing percentage of users. The 7-day expiration for first-party cookies written by external scripts is specific to Safari.
Do ad blockers also wipe out pixel data?
Yes. Blockers like uBlock Origin or Safari's built-in blocker block the call to the Meta pixel outright, so the event never fires. Server-side tracking isn't affected by blockers because it runs on your server, not in the user's browser.
Is there a difference between ITP on iPhone and on Mac?
Not a significant one. Safari on macOS implements the same version of ITP as Safari on iOS. On iPhone you can't change the browser engine — every browser on iOS uses WebKit, which applies the same tracking restrictions.
Can I see how many Safari conversions I'm losing in Meta?
Not directly, because those conversions arrive without identification data and Meta can't attribute them precisely. What you can see is your EMQ in Events Manager — if it's below 6, part of that is loss from Safari and ITP. An indirect indicator is comparing the percentage of Safari traffic on your site with the percentage of conversions Meta can identify in its event quality panel.
Does it make sense to use the browser pixel and server-side together?
Yes. The browser pixel is still useful for upper-funnel events (PageView, ViewContent, AddToCart) that happen before the purchase. Server-side complements it for purchase events, where identification data matters most. Use both with deduplication by event_id so you don't double-count conversions.
Keep reading
What Escalafy does for your store
Connect your channels and see the truth behind your numbers. No Excel, in real time.