Straight answer: you set up conversion tracking for Meta ads by installing the pixel, connecting the Conversions API alongside it (not instead of it), picking the right standard events, verifying your domain for Aggregated Event Measurement, and testing everything in the Test Events tool before you spend a pound on ads. Skip the Conversions API and you'll be optimising campaigns on data that's missing a meaningful chunk of your real conversions, and you won't even know it.
What you need in place before you touch Events Manager
Before any of the technical steps, you need four things sorted: admin access to Meta Business Manager, admin or editor access to whoever manages your website code (or the ability to add code yourself via Google Tag Manager), your business verified inside Meta Business Manager, and ownership or admin rights on the domain you're advertising to. That last one trips people up constantly. If your website sits on a domain your agency owns rather than you, you cannot verify it, and without verification you cannot configure Aggregated Event Measurement, which caps how many conversion events Meta can report from that domain.
Get the domain sorted first. Everything else builds on it.
Step 1: create your pixel and add the base code
Inside Meta Events Manager, click "Connect Data Sources," choose "Web," then "Meta Pixel." Name it after your business, not the campaign, you'll only ever need one pixel per business in almost every case. Meta will give you a base code snippet to add to every page of your site, inside the head tag.
Three ways to install it:
- Paste the base code manually into your site template if you built the site yourself or have a developer on hand
- Use Google Tag Manager, which is what I recommend for anyone running more than one platform's tracking, because it keeps all your tags in one container instead of scattered across template files
- Use a plugin if you're on WordPress and running WooCommerce or a page builder with a built-in Meta pixel field
If you're managing several campaigns and platforms at once, the right stack matters more than people admit. I've written before about the software options for running Facebook ad campaigns, and pixel installation is usually the first thing that goes wrong when someone's cobbled together three different tools without checking they talk to each other.
Step 2: set up the Conversions API, even if you think you don't need it
This is the step most small business owners skip, and it's the one that costs them the most money. The pixel is browser-side tracking. It relies on a cookie firing in someone's browser. Safari's Intelligent Tracking Prevention, ad blockers, and iOS 14.5's App Tracking Transparency prompt all interfere with that cookie firing correctly. Meta's own documentation admits pixel-only tracking under-reports conversions, and in my experience with client accounts the gap sits anywhere between 15% and 40% depending on how much of the traffic is on iPhone.
Conversions API sends the same event data server-side, straight from your website's server or your CRM to Meta, bypassing the browser entirely. Set it up alongside the pixel, not instead of it, and Meta will automatically deduplicate any event that both methods report, using an event ID you pass in both.
The setup itself depends on your platform:
- Shopify, WooCommerce and most major e-commerce platforms have a native Conversions API integration you switch on inside Events Manager in a few clicks
- If you're using Google Tag Manager's server-side container, you route events through that container to Meta's server endpoint
- For custom sites or lead gen forms, you can send events via webhook using a tool like Zapier or Make, which is also how a lot of people connect their lead ads to a CRM in the first place
If lead generation is your main goal rather than ecommerce, this is worth pairing with a proper lead ads setup. I've covered how Meta lead ads connect to your CRM through LeadsBridge, and getting that data flowing both directions (into your CRM and back to Meta as a conversion) is where the real optimisation gains sit, not just in the ad copy.
Step 3: choose the right events, not every event
Meta gives you a list of standard events: Purchase, Lead, CompleteRegistration, AddToCart, InitiateCheckout, Schedule, Contact, and more. Use standard events wherever one fits, because Meta's algorithm learns faster from standard events than from custom ones, and you'll need standard events to build lookalike audiences and run value-based bidding later.
For a lead gen business, fire the Lead event on the actual thank-you page after form submission, not on the form page itself. I see this mistake constantly, someone fires Lead on page load of the contact page and wonders why their cost per lead looks brilliant and their sales pipeline looks empty. The event should match the actual moment of conversion, nothing earlier.
For ecommerce, pass the value and currency parameters on the Purchase event every time. Without a value, Meta can't optimise for high-value customers, it just optimises for any purchase, and you end up with a full funnel of £8 orders instead of £80 ones.
Step 4: understand Aggregated Event Measurement and the 8-event limit
Once your domain is verified, Meta lets you configure up to eight conversion events per domain under Aggregated Event Measurement, ranked in priority order. This exists because of Apple's App Tracking Transparency rules, and it means that for users who've opted out of tracking, Meta will only report the single highest-priority event that fired in a session, not all of them.
So if you've got Purchase, AddToCart and ViewContent all configured, and a user views a product, adds to cart, then buys, only the Purchase event gets reported for that opted-out user, because it's ranked highest. Set your priority order to match what you care about, usually Purchase or Lead at the very top, then work down to softer signals. Most accounts I audit have this ordered wrong or haven't touched it since setup.
Step 5: test before you spend a penny
Inside Events Manager, open the "Test Events" tab and browse your own site with it open. Every event should appear in real time with a green tick and the correct parameters attached. If Purchase shows up with no value, fix it now, not after your first week of ad spend has trained the algorithm on bad data.
Also install the Meta Pixel Helper Chrome extension. It sits in your browser toolbar and flags duplicate pixel fires, missing parameters, and pixels that fired on the wrong page, in about ten seconds per page you check.
If you're running this alongside several other automations and tools for your campaigns, it's worth a wider check of your setup. I go through this in more detail in my piece on Facebook ad automation tools that save time on campaigns, because half the tracking errors I find come from tools firing conflicting events on the same page.
A real example, with real numbers
I worked with a home services client, boiler repair and installation, running lead ads and a landing page combo. When we first pulled their account, the pixel alone was reporting 38 leads a week at £22 cost per lead. Their CRM, which logged every phone call and form fill regardless of platform, showed 61 actual qualified leads coming from Meta traffic in that same week.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
That's a 38% gap, real money being spent on an optimisation signal that was missing well over a third of the actual outcome. We set up the Conversions API through a webhook from their CRM, deduplicated against the existing pixel events, and within three weeks Meta's reported lead number climbed to 57, still short of the CRM's 61 but close enough that we could trust the campaign data again. Cost per lead as Meta reported it went up slightly, to £26, because the algorithm now had visibility on leads it had previously been optimising blind to. The real cost per lead, calculated against the CRM number, dropped from roughly £34 to £24. The number on the dashboard got worse. The business got better. That's the bit nobody tells you when they're selling you a "10x your ROAS" course.
What breaks tracking (the mistakes I see on every audit)
- Firing the same event twice through two different tools without an event ID for deduplication, which inflates your reported conversions and makes campaigns look better than they are
- Tracking AddToCart as a custom conversion via URL matching on a page that also loads for people who abandon the cart afterwards, which counts intent as outcome
- Never checking the diagnostics tab in Events Manager, which flags issues like missing match quality on email or phone parameters, something that directly affects how well Meta can match your event to a real Facebook or Instagram user
- Setting up tracking once at launch and never revisiting it after a site redesign, a new checkout provider, or a CRM migration, all of which routinely break pixel placement without anyone noticing for months
None of this is exotic. It's mostly people setting tracking up once, under time pressure, and never going back to check it. If your ad account has been live for over a year and nobody's opened the diagnostics tab in that time, go and look today.
The uncomfortable bit nobody puts in the tutorials
Your Meta reported numbers and your Google Analytics numbers will never match, and chasing an exact match is a waste of a week you'll never get back. Meta uses a 7-day click and 1-day view attribution window by default, meaning it credits itself with a conversion if someone clicked an ad up to seven days before buying, or simply saw it and bought within a day without clicking at all. GA4 typically uses last-click or data-driven attribution, which credits the channel someone clicked through on right before converting. Two different measurement philosophies, two different answers, both defensible.
The businesses that get in a mess are the ones that treat Meta's dashboard as gospel truth and pause a campaign because ROAS "dropped," when their attribution window just changed how a normal sales cycle got counted that week. Use Meta's numbers to optimise inside the platform. Use your CRM or your bank account to judge whether the whole thing is working. Don't ask one tool to answer a question it was never built to answer.
Keeping it running once it's set up
Conversion tracking isn't a one-off job, it's a maintenance task, same as anything else in your account. Check Events Manager diagnostics monthly. Re-test after any website changes. And if you're scaling spend, revisit your event priority order every quarter, because what mattered when you launched (maybe just getting any Lead) often changes once you've got enough volume to optimise for higher-value ones instead.
If you're running campaigns across multiple ad accounts or brands, a decent management tool earns its keep here too. I've broken down what features matter in a Facebook ad management tool, and tracking visibility across accounts is one of the features people underrate until they're managing more than one.
And if this all sounds like more plumbing than marketing, that's because it is. It's also the plumbing that decides whether the rest of your campaign advice, targeting, creative, bidding, works. I've written a broader rundown of the fundamentals in tips for running Facebook advertising campaigns that hold up, and tracking sits underneath every one of them. Get it wrong and you're optimising a campaign against numbers that were never true in the first place. If you've tried to set this up twice and it still doesn't add up, that's usually a sign it needs an outside pair of hands rather than a third attempt, and that's exactly the kind of gap an AI consultant working with small businesses can close in an afternoon rather than another lost month.
Frequently asked questions
Do I need both the Meta pixel and the Conversions API, or just one?
Use both, together, with a matching event ID for deduplication. The pixel catches browser-side events, the Conversions API catches server-side events that the browser misses due to ad blockers, Safari's tracking prevention, or iOS privacy prompts. Running only the pixel typically under-reports real conversions by 15% to 40% depending on how much of your traffic is on iPhone.
Why don't my Meta conversion numbers match Google Analytics?
Different attribution windows. Meta defaults to a 7-day click and 1-day view window, meaning it can credit a conversion up to a week after someone clicked an ad, or the same day they simply saw it without clicking. GA4 generally credits the last channel someone clicked through. Both are correct by their own rules and will never line up exactly, so use Meta's numbers to optimise inside Meta, and your CRM or bank figures to judge overall performance.
What is Aggregated Event Measurement and why does it limit my events?
It's Meta's system for measuring conversions from users who've opted out of tracking under Apple's App Tracking Transparency rules. Once your domain is verified in Business Manager, you can configure up to eight priority-ranked events per domain. For opted-out users, only the single highest-priority event that fired gets reported, so put your most important conversion, usually Purchase or Lead, at the top of the list.
Where should I fire the Lead event, on the form page or the thank-you page?
On the thank-you page, after the form has been submitted, never on page load of the form itself. Firing it too early counts people who saw the form but never filled it in, which inflates your lead numbers and makes your cost per lead look far better than the real sales pipeline it's feeding.
Related reading: How Competitor Rank Tracking Works for SEO Strategy (And What It Won't Tell You) and What Is Competitor Tracking and Why You Should Use It for SEO.
For the bigger picture, see my full guide to digital marketing.
Related: shopify marketing strategy.
This goes deeper into Facebook ad tools.
