What Is Conversion Tracking? How It Works, Setup and Examples

Conversion tracking is the practice of recording when a visitor does something valuable on your site or app, such as buying, submitting a form, calling or signing up, and tying that action back to the ad, channel or page that brought them. A small piece of code records the action, and the data tells you which marketing produces results instead of just clicks.

This guide covers how conversion tracking works under the hood, what you can track, how Google Ads, Meta and GA4 each do it, how to set it up and test it, and why their numbers never match. It also covers what most guides skip: tracking conversions from channels that have no ad click, like organic search and AI assistants, and a worked example with real arithmetic.

What Is Conversion Tracking?

A conversion is a visitor completing the action you want. Conversion tracking is the system that notices each one and records where it came from. That second part is what makes it useful. In a hypothetical store, the order system knows you sold 120 things this month. Conversion tracking tells you that 45 of them came from one Google Ads campaign, 30 from organic search and 8 from an email.

The term is used in two slightly different ways. In ad platforms like Google Ads and Meta, “conversion tracking” means the feature that counts conversions after someone interacts with your ads, so you can report on them and let the platform bid toward them. In analytics, it means recording important actions from every traffic source, paid or not. Both rely on the same mechanics.

What Counts as a Conversion

Anything measurable that moves someone closer to paying you. Common examples, grouped the way ad platforms group them:

For a deeper list with counting rules for each, see our guide to conversion events.

How Conversion Tracking Works

Every setup, whatever the platform, follows the same three steps.

1. Remember where the visitor came from. When someone clicks an ad, the platform adds a click ID to the landing URL (Google uses gclid, Meta uses fbclid). A tag on your site stores it, usually in a first-party cookie. Analytics tools also read the referrer and any UTM parameters to work out the source.
2. Notice the action. A tag, also called a pixel or snippet, fires when the conversion happens: on the thank-you page that only loads after a purchase, or on the form submit or button click itself. It sends an event with details such as the value and an order ID.
3. Match the action to the source. The platform links the conversion to the stored click or visit and credits the ad, keyword, campaign or channel, according to its attribution rules and time window.

Google describes its own version plainly: you add the Google tag to your site, a cookie is set when someone clicks your ad, and when they complete the action the tag recognizes it and records a conversion. Some conversions need no tag at all, such as calls through Google forwarding numbers or app installs from Google Play (Google Ads Help, About conversion measurement).

Why Conversion Tracking Matters

• You see what pays, not what gets clicked. A campaign with lots of cheap clicks can produce nothing. Conversion data shows which campaigns, keywords and pages lead to revenue.
• You can move budget with evidence. Cost per conversion and return on ad spend let you cut what loses money and scale what earns it.
• Automated bidding needs it. Google and Meta bidding strategies that aim for conversions or a target cost learn from the conversions you send them. No data, or bad data, means the system optimizes toward the wrong thing.
• You find where people drop off. Tracking steps such as add to cart, checkout start and purchase shows the stage where a funnel leaks, which points to the page or form to fix.
• You can prove results. Agencies and in-house teams can report sales and leads, not traffic, which is what owners actually ask about.

The Metrics Conversion Tracking Produces

Once conversions are recorded, a handful of numbers follow directly:

An illustrative example: 2,000 clicks and 60 conversions is a 3% conversion rate. If those clicks cost $1,800, each conversion cost $30.

Conversion value is what turns these counts into money. Purchases carry their order value. Leads need a value you assign, based on how often they close and what a customer is worth. Our guide on how to calculate conversion value walks through the formula.

Conversion Tracking Methods Compared

There are several ways to record a conversion. Most accounts use two or three of them together.

If you want to go further on server-side setups, see server-side GTM.

Conversion Tracking in Google Ads, Meta and GA4

Google Ads

You create a conversion action under Goals, Conversions, choose its category (purchase, lead, and so on), how to count it (every conversion or one per click), its value, and its windows. The Google tag goes on every page, and the conversion is triggered either by a page URL or an event snippet. Google can also track calls and app installs, and accepts offline conversions uploaded with the click ID. The same Google Ads help article notes that click IDs used for offline imports must be within 90 days.

Meta (Facebook and Instagram)

The Meta Pixel loads on every page and sends standard events such as Lead, Purchase or Contact when they happen. The Conversions API sends the same events from your server, website platform or CRM. When you run both, give each event the same event ID so Meta counts it once.

Google Analytics 4

GA4 records events from all traffic, paid or not. You mark the events that matter as key events (GA4’s name for what it used to call conversions), and you can import them into Google Ads. Our GA4 conversion tracking guide has the full setup.

Conversion Tracking vs a Retargeting Pixel

The same base tag often does both jobs, which is why they get confused. A retargeting pixel observes behavior: pages viewed, products browsed. Platforms use it to build audiences you can show ads to later. Conversion tracking confirms an outcome: this person bought, this person submitted the form. One fills audiences, the other measures results. You need both if you run paid social or display, and having the pixel installed does not mean conversions are being counted until you define them.

Click-Through vs View-Through Conversions

A click-through conversion follows a click on your ad. A view-through conversion follows an ad impression with no click, within a set window. Search ads mostly drive click-through conversions. Display, video and social ads claim many view-through conversions, because people see an ad, do nothing, and buy days later.

Windows control how long after the click or view a conversion can still be credited. In Google Ads the click-through window defaults to 30 days and can be set from 1 to 90 days for Search and Display, the engaged-view window defaults to 3 days, and the view-through window defaults to 1 day (Google Ads Help, About conversion windows). Pick windows that match how long your customers actually take to decide; our guide to the attribution window compares platform defaults.

Treat view-through numbers with care. Someone who saw your ad and later bought may have bought anyway. The only way to know whether ads caused the extra sales is an incrementality test: hold some audiences or regions back from seeing ads and compare conversions.

Attribution: Which Touchpoint Gets the Credit

Buyers often touch several channels before converting. The attribution model decides how the credit is split:

• Last click: all credit to the final click. Simple, and it favors channels that close. More on last click attribution.
• First click: all credit to the touchpoint that started the journey.
• Linear, time decay and position based: fixed rules that spread credit across touchpoints. Google Ads and GA4 have retired these rule-based options, but many other tools still offer them.
• Data-driven: credit assigned from patterns in your own conversion paths. It is the default in Google Ads and GA4.

How to Set Up Conversion Tracking

The platform menus differ, but the work is the same everywhere.

Step 1: Decide what counts as a conversion

List the actions that make you money. Mark one or two as primary (purchase, qualified lead) and the rest as secondary. Bidding optimizes toward primary conversions, so a newsletter signup marked primary will pull spend toward cheap signups. Give each conversion a value.

Step 2: Install the base tag on every page

Add the Google tag, Meta Pixel or your analytics snippet site wide, either in the site’s head or through Google Tag Manager. The tag must be on the landing page to capture the click ID, not only on the thank-you page.

Step 3: Define the conversion

Use a URL rule if the action always loads a unique confirmation page. Use an event (form submit, button click, purchase event) if it does not, which is common with embedded forms and single-page sites.

Step 4: Send value and a unique ID

For purchases, pass the order total and the order or transaction ID. The ID lets platforms drop duplicates when a page reloads or the same event arrives from both the browser and the server.

Step 5: Test it

Complete the action yourself and confirm exactly one conversion is recorded with the right value. The next section shows how.

Step 6: Connect offline results and compare to your records

If sales close later by phone or with a sales team, save the click ID with each lead and import closed deals from your CRM. Then compare tracked totals with your order system or CRM every month.

How to Check Conversion Tracking Is Working

• Watch the tag fire. Use Google Tag Assistant, the Meta Pixel Helper or GA4 DebugView while you complete a real conversion on the live site, not only in preview mode.
• Read the status. In Google Ads, a new conversion action shows as Unverified until it records its first conversion or a successful test event. Data from GA4 links and offline imports can take 24 to 48 hours to appear (Google Ads Help).
• Count once. Reload the thank-you page. If a second conversion appears, add a transaction ID or switch to an event trigger.
• Reconcile with the backend. Compare a week of tracked conversions with your orders or CRM leads. Some loss is normal; a large or sudden gap means something broke.

Common Reasons Conversion Tracking Breaks

Google’s documentation names the payment processor case directly as a reason to list unwanted referrals (Analytics Help, Identify unwanted referrals). For consent settings, see consent mode v2.

Why Google Ads, Meta and GA4 Never Agree

Each ad platform counts the conversions it can link to its own ads, using its own windows and its own view-through rules. None of them knows about the others. When one buyer clicks a Google ad on Monday, sees a Meta ad on Wednesday and buys on Friday, both platforms claim the sale. GA4 sees the whole site and gives the sale to one channel.

So adding up platform reports will often give you more conversions than you actually had. That is not a bug in any one tool. It means platform numbers are for optimizing inside that platform, and a single neutral source (your analytics or your backend) is for comparing channels.

Worked Example: From Tracked Conversions to Real Returns

The numbers below are illustrative, for a small online store in one month, to show the arithmetic from start to finish.

What the example shows:

• The two ad platforms claim 70 + 55 = 125 conversions. The store only had 120 orders, and GA4 says 50 of them came from channels that were not ads at all.
• By its own report, Meta looks like the better buy (ROAS 2.75 vs 2.33). On the GA4 view it is the weaker one (1.25 vs 1.50). The ranking flips depending on which source you trust.
• At a 40% gross margin, a ROAS of 1.25 returns $1.25 × 40% = $0.50 of gross profit per $1 spent, so Meta loses money on the GA4 view, while the platform report says it is the star. That gap is the decision you need conversion tracking to make well.
• Organic search produced 30 orders ($3,000) with no ad spend, and neither ad platform reports any of it.

Conversion Tracking Without Ads: Organic Search and AI Assistants

Almost every guide to conversion tracking is about ads, because ads come with click IDs. Organic search and AI assistants like ChatGPT and Perplexity do not. There is no gclid on a Google organic result and no click ID on a cited link in an AI answer, so ad platforms cannot see these conversions at all. Only tools that see your whole site can.

• The source comes from the referrer. A visit from google.com is organic search; a visit from chatgpt.com or perplexity.ai is an AI assistant. When an app or browser strips the referrer, the visit lands in Direct, and no tool can honestly recover it.
• Keywords are hidden, so report by landing page. Search engines do not pass the search term, so the useful unit is the landing page: which organic pages start the sessions that convert, and what those conversions are worth.
• Value matters more than count. A blog post that brings 3 demo requests worth $600 each beats a page that brings 40 newsletter signups. Assign values and compare pages in dollars.

Our pillar guides cover both in depth: SEO conversion tracking and AI conversion tracking. If you want this without building it in GA4, SEOConversion is a cookieless, first-party tracker that records conversions from organic search and AI assistants and reports conversions and value by landing page, from one script.

Privacy, Cookies and Lost Conversions

Conversion tracking has always depended on remembering a visitor between the click and the action. Browser privacy features shorten how long cookies last, ad blockers stop tags from loading, and consent rules mean some visitors are never tracked. The result is that every browser-based setup misses some conversions.

Platforms respond in three ways: server-side sending (Conversions API, server-side Tag Manager), hashed customer data (Enhanced Conversions), and modeling of the conversions they cannot observe. Each recovers some signal; none makes the numbers exact. If you want to measure without third-party or persistent cookies, see cookieless tracking.

FAQ