What Is Ad Tracking? A Beginner’s Guide to Measuring Your Ads
Learn how ad tracking works, what it measures, and how to test your setup with a simple example.
Ad tracking is the collection of information about what happens around an ad, such as how often it appears, whether someone clicks it, and whether a useful action follows. It helps you read campaign reports and investigate what may need attention. What a tool can observe depends on its setup and the visitor's choices.
What is ad tracking?
An impression is a recorded display of an ad. A click is an interaction that takes someone to a destination. A conversion is an action you want to measure, such as a purchase or a signup. Ad tracking brings information about these events together so you can compare campaigns.
For example, imagine a fictional online plant shop. It runs an ad for a starter kit and wants to know how many visitors reach the product page and later buy. The shop needs a way to recognize the campaign on arrival and a separate way to record the purchase.
How ad tracking works, step by step
- The ad appears. The ad platform records an impression according to its own measurement rules.
- Someone clicks. The visitor opens the landing page, the page they reach after clicking. The link may include labels called UTM parameters or a click identifier that helps a tool recognize the source.
- The shop records an action. A small piece of website code, often called a tag or tracking pixel, can report an event such as a product view or checkout start. A completed order may also be sent from the store's backend, the system that processes the order.
- A report assigns credit. Attribution is the set of rules a tool uses to assign conversion credit to observed marketing interactions. It may give credit to a recent click or distribute it across several interactions.
In our fictional example, a visitor clicks the plant ad, browses, then returns later through an email and buys. One report might credit the ad; another might credit the email. The result depends on each tool's rules and the information it received.
What can you measure?
| Metric | Meaning | Simple calculation |
|---|---|---|
| Impressions | Recorded ad displays | Counted by the ad platform |
| Clicks | Recorded clicks | Counted by the ad platform |
| Conversions | Measured desired actions, such as completed orders | Counted by the configured system |
| Conversion rate | Conversions relative to a chosen base | Conversions ÷ clicks × 100 |
| Cost per acquisition (CPA) | Ad spend per attributed acquisition | Ad spend ÷ attributed acquisitions |
| Return on ad spend (ROAS) | Attributed revenue compared with ad spend | Attributed revenue ÷ ad spend |
Fictional example: suppose the plant shop spends $200, receives 100 clicks and attributes 5 purchases worth $500 in total revenue to the campaign. Using clicks as the base, its conversion rate is 5%, CPA is $40 and ROAS is 2.5, or $2.50 in attributed revenue for each $1 spent on ads.
ROAS is not profit: it does not subtract product, shipping or other business costs. Check which conversions, revenue and time period each report uses before comparing figures. If clicks or conversions are zero, the corresponding division cannot be calculated.
Common ways to track an ad
UTM parameters are labels added to a URL, such as source and campaign name. They provide context for a visit; they do not record a purchase by themselves. Google's URL guide explains how campaign parameters work. For a practical setup, see our Meta Ads UTM guide.
Click identifiers are values carried with a click that can help a measurement system match a visit to an ad. Their meaning depends on the platform and integration. Clakta describes its separate clkt key in Tag your ad URLs.
Pixels and tags are small pieces of code that can report selected activity from a website. They need to be configured for the events you want to measure and tested with your consent settings.
Server-side tracking processes or sends events through a server. A browser can still be the source of an initial event; the server is another part of the route, not a guarantee that all events will be captured. Google's introduction explains one server-container approach, and our server-side tracking guide covers the tradeoffs.
Cookies and other browser storage can retain information across page loads or visits, depending on configuration. They have a different role from URL labels and event-sending tags; neither is a synonym for a cookie.
How to get started
- Pick a goal you can define clearly, such as a completed order or a submitted lead form.
- Give campaign links consistent labels and follow the instructions for your measurement tool.
- Configure the events that represent your goal. Decide when an order counts as complete.
- Open a test link, check that its parameters survive redirects, and perform a test action. Compare the event with the relevant order or lead record.
- Review reports over the same dates and settings. Investigate missing or duplicate events before changing a budget.
If you use Clakta, its tracking overview describes how the visit context and an order from a connected store or backend are brought together. Follow the setup instructions for your own integration.
Why numbers differ and what tracking cannot tell you
Platforms can use different attribution models and attribution windows, the time periods in which an interaction is eligible for credit. A purchase can be reported by more than one platform, so adding platform totals can double-count it. Different event definitions and reporting dates can also change a comparison.
Some events or identifiers may be missing because of consent choices, browser behavior, failed requests or a change of device. Reports are useful, but an attributed conversion does not prove the ad caused an extra sale. An incrementality experiment asks a separate question: what changed because people were exposed to the ad?
Privacy and consent
Privacy requirements depend on where people are located, the data involved and the implementation. Decide what data is needed, explain its use and make sure the setup respects applicable consent choices. Using a first-party domain or a server does not automatically remove consent requirements. Google's consent overview illustrates how its tags can respond to consent signals; it does not determine your legal obligations.
Frequently asked questions
How does ad tracking work?
A platform records ad activity, and a configured website or backend records selected actions. Reporting tools use available context and their attribution rules to connect and summarize the observations.
Is ad tracking the same as attribution?
No. Tracking collects or records observations; attribution assigns credit among eligible interactions. Attribution depends on what the tool can observe.
Do UTMs track conversions by themselves?
No. UTMs label a visit. A separate event or order record is needed to measure the conversion, and the system must retain enough context to connect it to the visit.
Can you track without cookies?
Some forms of measurement can operate without browser cookies, but the available data and ability to connect visits can be different. Whether a particular approach is appropriate depends on its configuration and consent requirements.
Is ad tracking legal?
That depends on the jurisdiction and how the data is collected, used and shared. Review the rules that apply to your business and obtain qualified advice when needed; a tracking technology alone does not establish compliance.
Start with one campaign and one clearly defined conversion. Test the path from the tagged link to a real record before relying on a report to make decisions.