Meta Conversions API (CAPI) and Your First-Party Data Strategy: A 2026 Guide for Shopify

The Meta Conversions API (CAPI) is a server-to-server connection that sends conversion events, like purchases and add-to-carts, from your backend directly to Meta instead of relying only on the browser-based Meta Pixel. It is the main way Shopify brands recover the signal lost to ad blockers, iOS privacy changes and browser tracking limits, and it works best when it is fed with high-quality first-party data.

This guide explains how Meta (Facebook, Instagram, Messenger and WhatsApp) and other ad platforms use first-party data. It covers the technical details of the Meta Pixel and the Meta Conversions API (CAPI), formerly called the Facebook Conversions API, and how Shopify brands use both for retargeting, dynamic product ads and lookalike audiences. It ends with a seven-step first-party data strategy for Meta in 2026, in the order that pays off fastest. Throughout, we've added what we see when auditing Meta setups for Shopify brands every week.

What is first-party data?

First-party data refers to information a business collects directly from its own sources, such as website visitors, app users, email subscribers or in-store purchase records. Because it's gathered from direct interactions, first-party data is more accurate, privacy-compliant, and ethically sourced than third-party data (which is often collected by external brokers).

For a Shopify brand, first-party data typically includes:

• Behavioral data: page views, product views, add-to-carts and checkouts on your own store.
• Customer identifiers: email addresses, phone numbers and shipping addresses captured at checkout, account creation or a signup form.
• Transaction data: orders, revenue, refunds, subscriptions and in-store purchases.
• Engagement data: email and SMS clicks and opt-ins, quiz answers and support conversations.

---

Why is first-party data important for Meta ads?

Meta is not an analytics platform; it is a prediction engine. It doesn't just need to know how many people added to cart, it needs to know who they were, so it can find more people like them and decide who should and shouldn't see your ads. The more of your conversions Meta can tie to a real person, the better it targets. First-party data is the most reliable source of that signal for four reasons:

• Accuracy: Directly collected user interactions (e.g., page visits, add-to-cart actions) reflect actual customer behavior.
• Privacy compliance: Safari and Firefox already block third-party cookies by default, and while Google dropped its plan to phase them out of Chrome, ad blockers, Safari's Intelligent Tracking Prevention and iOS privacy features still erode browser-based tracking. First-party data collected with consent remains a reliable alternative.
• Campaign effectiveness: First-party data fuels better audience targeting, dynamic product ads, and retargeting campaigns.
• Reduced data loss: By owning the data pipeline, advertisers mitigate the impact of ad blockers, iOS tracking restrictions, and browser-level tracking prevention.

Since Apple's iOS 14 changes, much of the matching Meta used to do on its own has shifted to advertisers: Meta now relies on you to send it complete, well-identified conversion data. Poor data shows up in recognizable ways: ads that look strong on click-through rate but stop getting spend, or results that swing from great to bad week to week. For a deeper look at what incomplete data costs, see The True Cost of Bad Marketing Data.

---

How does the Meta Pixel collect data?

The Meta Pixel (or "Facebook Pixel") is a snippet of JavaScript code placed on your website. It effectively tags visitors' actions (events) and sends these signals back to Meta's advertising platform.

Typical events captured by the Meta Pixel

• PageView: Triggered when someone visits a page on your website.
• ViewContent: Fired when someone views a product detail page.
• AddToCart: Logged when a user adds a product to their cart.
• InitiateCheckout: Captured upon starting the checkout process.
• Purchase: Fired once a transaction is completed.

Data flow

1. User lands on a webpage.
2. Pixel code collects relevant event data (e.g., product ID, price, currency).
3. Pixel sends this data to Meta, where it is processed and attributed to a specific user or device if possible.

Technical note: The Pixel relies primarily on first-party cookies (_fbp for the browser and _fbc for the ad click) and browser-local storage for session tracking, after iOS and browser privacy restrictions significantly impacted third-party cookies.

Advanced Matching: If a user logs in or enters an email address at checkout, that hashed PII can be sent to Meta for more accurate matching.

---

Why does the Meta Pixel miss conversions?

The Meta Pixel misses conversions because modern shopping journeys break across browsers, devices and sessions, and the browser can't connect them. When the connection breaks, the ad click ID is lost and Meta can no longer tie the purchase to the ad that drove it.

A typical Shopify journey looks like this:

1. A shopper taps an Instagram ad and browses inside Instagram's in-app browser.
2. They don't buy there. In our experience, most shoppers don't complete a purchase in the in-app browser. Instead they open the store later in Safari, or on a laptop.
3. They come back days later through a Google search or an email and finally buy.

To the browser, that's three different anonymous visitors. The click ID from the original ad lives in the first session only, so the purchase arrives without it, and Meta has to fall back on matching by email, if an email is attached at all. Our guide to fbclid and the _fbc cookie explains how the click ID is stored and where it gets lost. Several other things make it worse:

• Safari cookie limits: Safari caps cookies set by JavaScript at seven days, so a returning customer is often treated as a brand-new anonymous visitor.
• Ad blockers: blocked browser events never reach Meta at all. Tag-manager-based setups are frequently blocked too.
• Sandboxed checkouts: InitiateCheckout often under-fires in Shopify's sandboxed checkout and with express wallets like Shop Pay and PayPal, so it needs a server-side backup.
• Theme and app changes: a new theme launch can silently switch off a tracking app embed, and browser events quietly stop.

The result is under-reported conversions in Ads Manager and a smaller, less accurate picture of who your buyers are.

---

What is the Meta Conversions API (CAPI)?

The Conversions API (CAPI) is a server-to-server data pipeline that allows you to send user events from your backend systems directly to Meta's servers. Instead of relying solely on the client-side Pixel, which can be blocked by ad blockers or restricted by iOS privacy measures, CAPI ensures a more reliable and complete event flow.

Server-side tracking means sending the data from a server instead of from the shopper's browser. The events still go to the same Meta Pixel (dataset) in Events Manager; only the delivery route changes. You'll see it called several names (Meta Conversions API, Facebook Conversions API, Meta CAPI or Facebook CAPI), and they all refer to the same thing. Meta's official reference is its Conversions API documentation.

Key benefits

• Resilient to browser restrictions: Data is sent from your server, bypassing many client-side limitations.
• Improved data accuracy: Reduces data loss from ad blockers or network issues on the client's side.
• De-duplication: Meta can de-duplicate events so that a single purchase isn't counted twice if both Pixel and CAPI fire (see how deduplication works).
• Better attribution: More events received means more complete funnel insight, improving how Meta optimizes your campaigns.
• Richer signal: Your server can attach data the browser never sees, such as new-versus-returning customer status, subscription details or customer identifiers from your email platform.

---

How does the Conversions API work?

When a customer converts, your server sends the event to Meta over an authenticated API call, and Meta matches it to a person using the customer identifiers attached to it.

1. Server-side event generation: When a user purchases a product on your Shopify store, your backend logs that event.
2. API call to Meta: Shopify (or another server) sends a POST request to Meta's Conversions API endpoint, containing event data (e.g., purchase amount, product ID, user email hashed).
3. Data matching: Meta uses identifiers (like hashed email, phone number, or other user data) to match events to user profiles, which improves attribution accuracy.

How well step 3 works is measured by Meta's Event Match Quality (EMQ) score. Despite the name, EMQ is really an audience match score: it tells you how often Meta can tie your events to a real Facebook or Instagram user. Our complete guide to Event Match Quality explains what drives it.

---

What data does the Conversions API send to Meta?

Each Conversions API event contains three kinds of data: the event itself, customer information used for matching, and custom data describing what happened. Events can be sent up to seven days after they occur.

Standard events

CAPI supports the same standard events as the Pixel. For e-commerce, the most important are ViewContent, AddToCart, InitiateCheckout, AddPaymentInfo and Purchase. Lead-generation and subscription brands also commonly send Lead, CompleteRegistration and Subscribe. Don't neglect the upper funnel: Meta uses PageView, ViewContent, AddToCart and InitiateCheckout to build intent audiences and decide who gets retargeted.

Customer information parameters

Meta's customer information parameters are the identifiers it uses to match an event to a person. Personal fields must be normalized and SHA-256 hashed before sending.

Which identifiers matter most?

• Click ID (fbc): the most valuable single signal, because it tells Meta three things at once: which person, which ad and when they clicked. See what fbclid is and how fbc is formatted.
• Email: stable for years, which makes it a strong match key.
• Phone: useful, but numbers get recycled, so it's weaker than email.
• External ID: doesn't map to a Facebook profile on its own, but lets Meta recognize the same person across visits.
• IP address and user agent: supporting signals only; IP addresses are no longer reliable enough to identify someone by themselves, and names add little.

Custom data

Custom data describes the conversion itself: value, currency, content_ids, contents and num_items. Value and currency are required for Purchase events, and content_ids must match your catalog for dynamic product ads to work.

---

Meta Pixel vs Conversions API: what's the difference?

The Pixel tracks events in the visitor's browser; the Conversions API sends events from your server. They capture the same actions, but they fail in different ways, which is why Meta recommends running both.

How a single shopping session flows through both

• User activity: The user visits a Shopify store, browses products, and adds items to their cart.
• Pixel tracking (browser-side): JavaScript events capture these actions and send them to Meta. Prone to ad-blocker and privacy restrictions (e.g., iOS 14.5+ "App Tracking Transparency," Safari ITP).
• Conversions API (server-side): Shopify's backend logs the same events or gathers them at checkout completion. Sends them securely to Meta's servers via an HTTP POST request, bypassing many browser limitations.
• Data matching and de-duplication: Meta identifies if an event from Pixel and an event from CAPI represent the same user action, marking duplicates accordingly. This ensures accurate reporting and optimization.
• Campaign optimization: Meta's machine learning algorithms use these events to refine targeting, optimize bidding, and serve more relevant ads.

---

Do you need both the Pixel and CAPI?

Yes. Meta's recommended setup sends every event from both the browser and the server, and Meta deduplicates them. In Events Manager, each key event should show both "Browser" and "Server" as connection methods. An AddToCart or ViewContent event that arrives from the browser only is a red flag that you're losing signal.

The catch is that the redundancy only works with exactly one source per channel and correct deduplication. Otherwise, purchases are counted twice and your reported CPA looks better than it really is.

---

How does Meta deduplicate Pixel and CAPI events?

Meta deduplicates a browser event and a server event when both share the same event_name and the same event_id and arrive within 48 hours of each other. Meta keeps one and discards the other (Meta's deduplication guide).

• Generate one ID per action: create a unique event ID when the action happens, and pass the same value to both the Pixel (eventID) and CAPI (event_id).
• Match event names exactly: Purchase and purchase are treated as different events.
• Use one sender per pixel: Meta deduplicates well when browser and server events come from the same source, and poorly across different tools, because each tool generates its own event IDs. Our event ID deduplication guide shows the code on both sides and which Shopify ID to use.

Why running two tracking tools double-counts

The most common problem we find in audits is duplicate data, not missing data. Typical causes on Shopify:

• Shopify's Facebook & Instagram app still sharing data alongside another server-side tool. Keep the app installed (you still need it for your catalog and Shops), but turn off its data sharing so only one tool sends events. This setting is often switched back on by a new team member, so it's the first thing to check.
• Leftover tags from a previous tool, such as old Google Tag Manager containers or a forgotten Conversions API Gateway connection, still firing.
• Off-Shopify landing page snippets pasted into the theme, firing a second pixel.

Meta's fallback compares fbp or external_id, and only when the browser copy arrives first; it never pairs two server events, so it can't rescue purchases, PageView, ViewContent or AddToCart sent by two server-side tools. Inflated upper-funnel events make Meta think buyers are coming, so it bids up for traffic that doesn't convert. Duplicates that slip through are often credited once to a click and once to a view, inflating view-through conversions. If Meta ever reports more purchases than Shopify recorded for the same day, time zone and attribution window, you have a duplicate source. Our guide to Meta attribution pitfalls and event duplication walks through more examples.

Upstack sends browser and server events from a single install, deduplicated on a shared event ID, so higher match quality doesn't come with double-counted purchases.

How do Shopify brands use first-party data with Meta?

Shopify merchants frequently integrate their stores with Meta Pixel and Conversions API to collect and use first-party data. Common use cases:

Dynamic product ads

• How it works: Shopify merchants generate a product catalog in Meta Commerce Manager. The Pixel or CAPI collects product IDs for each event (ViewContent, AddToCart, etc.).
• Value: Meta dynamically displays the exact products a user viewed, added to cart, or similar items, which tends to lift conversion rates.

Retargeting and remarketing

• Pixel tracking: When visitors leave without purchasing, the Pixel logs their activities.
• CAPI backup: If Pixel tracking is blocked or incomplete, CAPI ensures server-side event capture.
• Outcome: Merchants can create retargeting campaigns to re-engage these visitors, offering incentives or reminding them of abandoned carts.
• Building the audiences: our guide to Facebook custom audiences from server events and first-party data covers which audiences to build and why they come out smaller than your traffic.

Lookalike audiences and exclusions

• First-party data source: Merchants compile a high-value customer segment (e.g., top 1,000 spenders).
• Upload/sync: This segment becomes a Custom Audience in Meta, either uploaded as a customer list or built from Pixel and CAPI events. Advanced matching and rich CAPI identifiers raise the share of customers Meta can match.
• Lookalikes: Meta finds users who share demographics, interests, and behaviors with that seed audience, expanding your audience reach beyond existing customers. Manual lookalikes are used less often today, but Meta's AI targeting applies the same idea automatically: it looks for people like the buyers it can identify, which is why match quality matters even if you don't build lookalikes yourself.
• Exclusions: Better-identified customers also make exclusions work. Brands are often surprised how many "new customer" purchases turn out to be returning customers once they can measure it.

Upstack's predictive audiences for Meta automate this by syncing your highest-value customer segments as they change.

New-customer and subscription conversions

• New-customer purchases: sending a custom "new customer purchase" event, checked against your full Shopify order history, lets you add a custom conversion column in Ads Manager and see new-customer CPA and ROAS directly.
• Subscriptions: subscription brands benefit from a separate first-subscription event, and should not send recurring renewal orders as Purchase events, which teaches Meta that renewals are new conversions.
• What to optimize on: most brands still get the best results optimizing campaigns on the standard Purchase event, using new-customer and subscription events for measurement. Narrowing optimization to a smaller event shrinks the pool Meta can learn from.

Offline conversions

• In-store purchases: If a Shopify merchant also has a physical store, offline purchase data can be sent to Meta using CAPI. Meta discontinued its separate Offline Conversions API in May 2025, so the Conversions API is now the way to send offline events.
• Omnichannel attribution: This ensures that if someone saw an online ad and later bought in-store, the event is attributed appropriately.

---

How do you set up the Conversions API on Shopify?

There are four common ways to send Conversions API events from a Shopify store. They differ mainly in how much control you get and how much identity data reaches Meta. Whichever you choose, use only one of them per pixel.

With Upstack Signal, every conversion reaches Meta with hashed email and phone, a restored click ID and a stable external ID. Signal stitches sessions across browsers and devices, including visitors who return from a Klaviyo email, against a network of more than 750 million US device IDs, so the Instagram-to-Safari-to-laptop journey reaches Meta as one person. Setup through the Shopify integration takes under 20 minutes. If you're comparing tools, see Upstack vs Elevar and Upstack vs Aimerce, and for a broader checklist, our Shopify tracking setup guide.

---

How do you check if the Conversions API is working?

Open Meta Events Manager, select your dataset (Pixel), and confirm that your key events show "Server" as a connection method alongside "Browser." Then work through these checks:

1. Browser and server lines move together. For AddToCart and Purchase, browser and server event counts should track each other. Small gaps are normal, and the server line is often a little higher because it catches events the browser misses; one line at double or half the other means a missing or duplicate source.
2. Compare purchases with Shopify. Using yesterday's data (not today's, which is still incomplete), the same time zone and the same attribution window, Meta should never report more purchases than Shopify recorded.
3. Count your pixels. Use your browser's network tab or Meta's Chrome extension (Meta Pixel Helper, now called Meta Ads Data Advisor) to see which pixel IDs load on each page. More than one of your own pixel IDs, or the same event arriving twice with different event IDs, points to a second source. Event IDs beginning with sh- come from Shopify's Facebook & Instagram app. Our Meta Pixel Helper guide walks through each check.
4. Check click ID coverage. The share of events carrying a click ID should roughly match the share of your traffic that comes from Meta ads. If Meta drives most of your traffic but only a small fraction of events carry a click ID, clicks are being lost. Brands with heavy organic or affiliate traffic will naturally have lower coverage.
5. Review Event Match Quality. Check EMQ per event, especially Purchase. Our step-by-step guide to improving EMQ covers how to raise it.
6. Use Test Events and Diagnostics. The Test Events tab shows server events arriving in real time when you place a test order, and Diagnostics flags missing parameters, deduplication problems and delayed events.

Give changes time: data should be judged on full days, EMQ typically starts improving within 48–72 hours of a fix, and it usually settles within about two weeks.

---

Is the Conversions API mandatory for Meta ads?

No. You can run Meta ads with only the Pixel, and Meta does not require the Conversions API. However, Meta strongly recommends it, and browser-only tracking misses a meaningful share of conversions. Because Meta optimizes toward the conversions it can see, missing conversions mean higher reported CPAs, weaker targeting and less stable results. Campaigns using cost caps feel it most: if Meta can't attribute a sale, the cap can't work.

---

What is the Conversions API Gateway?

The Conversions API Gateway (CAPIG) is Meta's own tool for setting up the Conversions API without writing code. It runs on a cloud server, either in your own cloud account or hosted by a partner, and automatically sends a server-side copy of the events your Pixel already tracks.

It's quick to deploy, but it only forwards what the Pixel captures. It doesn't add customer identifiers the browser never saw, send order-level data such as new-customer status, or recover purchases the Pixel missed entirely. It's also easy to forget about: a Gateway left connected after switching tools will keep sending duplicate events alongside your new setup. Our guide to the Conversions API Gateway compares it with a direct integration, server-side GTM and managed options.

---

How to build a first-party data strategy for Meta in 2026

A first-party data strategy for Meta means collecting customer data you own, sending it to Meta in the most matchable form, and using it to shape who Meta targets. Fix tracking before you invest in reporting: dashboards only describe performance after the fact, while the data you send Meta shapes performance itself.

What changed for first-party data in 2026

• Meta decides more of the targeting. Advantage+ campaigns and broad audiences leave fewer manual levers, so the conversion data you send through the Pixel and Conversions API is the main input you still control.
• Browser tracking keeps eroding. Google kept third-party cookies in Chrome, but Safari's seven-day cap on script-set cookies, ad blockers and iOS privacy features still break browser-only tracking, so server-side events and stored click IDs matter more each year.
• Health data is categorized, not renamed. Since January 2026, Meta restricts health-related data by what it contains rather than the event name, so custom-event workarounds no longer restore optimization (details below).

These seven steps cover the strategy, in the order that delivers the biggest gains first:

1. Get every event to Meta from both browser and server. Start with Purchase, then make sure InitiateCheckout, AddToCart and ViewContent are server-backed too.
2. Remove duplicate sources and deduplicate. One sender per pixel, with a shared event_id between browser and server events (how it works).
3. Enable advanced matching and enrich identifiers. Capture hashed emails or phone numbers at checkout, account creation or signup, and attach them to every event you can, not just Purchase. Identity resolution goes further by recognizing returning visitors and attaching known identifiers to anonymous sessions, which is how Upstack Signal raises match quality on upper-funnel events.
4. Preserve click IDs across sessions and devices. Capture fbclid from landing URLs and the _fbc and _fbp cookies, store them server-side, and send them with every event, so Safari's cookie limits and device switches don't break attribution.
5. Handle consent correctly. Display clear consent banners and adhere to regulations (GDPR, CCPA) when collecting personal data, and only send events and identifiers your customers have agreed to share.
6. Monitor Event Match Quality and click ID coverage. Check EMQ per event and the share of events carrying a click ID, and fix gaps (checklist).
7. Connect your other first-party data. Use email and SMS engagement from Klaviyo as additional identity and intent signals, add new-customer and subscription events for measurement, and sync high-value segments back to Meta as Custom Audiences and exclusions.

A note for health and wellness brands

Meta restricts the data it will use from domains and datasets it classifies as health-related. When that restriction applies, lower-funnel events like AddToCart, InitiateCheckout and Purchase can effectively stop being used for optimization, and appeals rarely succeed. The current remedy is a new, clean dataset that receives server-side events only, sent from a neutral domain with product names, URLs and health terms removed, and optimized on the standard Purchase event rather than a renamed custom event.

Custom-event workarounds stopped working in January 2026, because Meta categorizes the data rather than the event name. See how to get purchase optimization back after a health & wellness flag, and our guide to Meta's health and wellness restrictions for what Meta restricts and how to check your dataset.

---

How do Google and TikTok use first-party data?

While Meta is a major example, Google Ads, TikTok Ads, Snapchat Ads, and others also have server-side APIs or "events" frameworks. They similarly combine client-side tracking tags (e.g., Google Tag, TikTok Pixel) with server-side signals. The overarching goal remains the same: deliver more accurate event data for better ad targeting and measurement. The same rule applies on every platform: one tracking source per destination, or conversions get counted twice.

• Google Ads: Server-side tagging and enhanced conversions send conversion data from a server container, avoiding some browser-level blocks. See the Google Ads conversions integration.
• TikTok: The TikTok Events API pairs the TikTok Pixel with server events to support retargeting and lookalike audiences. See the TikTok Events API integration.

---

Frequently asked questions

What is a first-party data strategy for Meta?

It's a plan for collecting customer data you own, such as emails, phone numbers, click IDs and order history, and sending it to Meta through both the Pixel and the Conversions API so Meta can match conversions to real people. In 2026 that means server-backed events, one deduplicated source per pixel, hashed identifiers on as many events as possible, stored click IDs and consent handled correctly.

What is a good Event Match Quality score?

Meta rates EMQ from 1 to 10 and recommends a score of 6.0 or higher. Across the Shopify accounts we work with, Purchase typically lands between about 8.9 and 9.3, because shipping orders include full customer details; below 8.5 is low. Upper-funnel events score lower because visitors haven't shared contact details yet: PageView around 6.2–6.5 is on the low end, and roughly 7–7.5 is healthy. Treat EMQ like a check-engine light: a low score can hurt performance, but a high score doesn't cause good performance on its own. See our EMQ guide.

Is the Conversions API GDPR compliant?

The Conversions API can be used in a GDPR-compliant way, but compliance depends on your setup rather than the API itself. Only send events and identifiers for visitors who have consented, hash personal data before sending it, and document Meta as a data processor in your privacy policy. For US state privacy laws, Meta also offers a Limited Data Use setting.

Does the Conversions API fix iOS 14+ tracking loss?

Partly. CAPI isn't blocked by browsers or by Apple's App Tracking Transparency prompt, so it recovers conversions the Pixel misses. It can't restore tracking of activity inside the Facebook and Instagram apps for users who opted out, but it gives Meta far more complete conversion data to match and optimize on.

Does Shopify's native Meta integration include the Conversions API?

Yes. Shopify's Facebook & Instagram app sends Conversions API events when data sharing is turned on. It's a reasonable starting point, but it offers limited control over which identifiers are sent and how events are deduplicated. If you switch to another server-side tool, keep the app installed for your catalog but turn its data sharing off (why).

Why does Meta show more purchases than Shopify?

Usually because two tools are sending events to the same pixel, most often Shopify's native data sharing plus a second server-side tool, or leftover tags from a previous setup. Different tools use different event IDs, so Meta can't fully deduplicate them. First confirm you're comparing the same day, time zone and attribution window, then look for the duplicate source (checklist).

How much does the Conversions API cost?

Meta doesn't charge for the Conversions API itself. Costs come from how you run it: server hosting for the Gateway or server-side GTM, or a subscription for a managed partner. Upstack plans start at $99 per month; see pricing.

---

The future of first-party data and CAPI

The advertising ecosystem is moving away from third-party cookie reliance. First-party data has become the cornerstone of marketing strategies, especially for Shopify merchants who want to build direct customer relationships and maintain reliable, privacy-compliant tracking.

Running the Meta Pixel and the Conversions API together captures more complete data and keeps your tracking resilient to browser and platform privacy changes. Paired with good analytics and CRM data, first-party data helps merchants:

• Deliver highly personalized product recommendations
• Execute effective retargeting for abandoned carts
• Build precise lookalike audiences for scaling
• Gain deeper omnichannel insights (online + offline purchases)

As privacy rules and consumer expectations change, the brands that collect first-party data with consent and send it to ad platforms reliably, through methods like server-side tracking, will have the edge. Don't let Meta guess when it doesn't have to.

Want to see how much signal you're leaving on the table? Request a demo and we'll audit your current Meta setup, or start a free trial and compare your Event Match Quality before and after. See results from other brands in our case studies.