Tracking is how you prove a conversion was yours — and get paid for it. Weak tracking means lost money you cannot even see, because the sales still happen; they just never get matched back to the click you paid for.
Everything else in performance marketing sits on top of this one foundation. Before you worry about creatives, offers or scaling, you need a clean chain that carries a single identifier from the moment someone clicks all the way to the moment they convert. This guide walks that chain end to end, explains the two ways a conversion can be reported, and shows why disciplined operators treat server-to-server tracking as the default rather than the upgrade. If the surrounding vocabulary is new, keep the analytics for beginners primer open alongside this.
The journey has four moving parts, and each one has to hand off cleanly to the next. When a user clicks your link, the tracker mints a unique click ID and stamps that one click with it. That ID then follows the user through to the offer, riding along in the URL so it survives every redirect between your link and the advertiser's page. When the user completes the action — a sale, a lead, a signup — the advertiser or network fires a postback that carries the same click ID back to the platform that issued it. Finally the platform reads that ID, matches it to the original click, and credits the conversion to you with its payout and status. Break any link in that chain and the money still moves, but it moves without your name on it.
The click ID (some networks call it a subid, cid or tid) is the single value that matters most. It is a random string that uniquely identifies one click and ties it to a campaign, a source and an offer. Everything downstream depends on that thread surviving intact: if it is dropped by a landing-page redirect, overwritten by a second tool, or returned in a field your tracker is not reading, the conversion simply will not match. This is why operators obsess over passing the ID correctly before they ever touch bids or creative — a broken click ID is invisible in the ad account and only shows up as a gap between what you know you drove and what the network paid.
There are two mechanisms for telling the platform a conversion happened, and they fail in very different ways. A pixel fires inside the user's browser and leans on cookies; a postback fires between servers and carries the click ID in the URL, touching no browser at all. The comparison below is the whole argument in one view.
| Method | Where it fires | Breaks when | Best for |
|---|---|---|---|
| Browser pixel | In the user's browser | Ad blockers, cookie loss, JS errors | Retargeting & on-page signal |
| Server-to-server (S2S) | Server to server | Rarely — no browser in the path | The confirmed conversion that pays you |
Most seasoned operators run both — S2S as the source of truth for what they get paid on, and a pixel layered on top purely for the retargeting audiences and behavioural data a server call cannot capture.
Browser pixels break in ways you never see. Ad blockers strip them, privacy settings block them, and expired or isolated cookies quietly erase the link between click and conversion. Server-to-server (S2S) postbacks fire directly between the advertiser's server and the tracker, so they do not depend on the user's browser, survive cross-device journeys where a cookie could never follow, and are far harder to defraud because they tie to a confirmed backend event. That is why disciplined operators run S2S with clean click IDs and treat pixels as the optional extra. The full mechanics — macros, the postback URL, testing and deduplication — are covered in postback & S2S tracking.
The ground under browser tracking has moved, and it is browser policy — not your setup — driving it. Safari and Firefox already block third-party cookies by default and isolate first-party ones, so a large slice of your traffic is effectively cookieless regardless of what you do. Chrome is the twist worth getting right: after years of announcing that it would deprecate third-party cookies, Google reversed the plan, so third-party cookies remain enabled by default in Chrome as of 2026 — Chrome is now the only major browser where that is still true. The honest takeaway is not "cookies are going away," but "do not count on them," because so much of your audience is already cookieless. Server-side S2S sidesteps the whole mess by never reading or writing a third-party cookie at all.
Most tracking problems are not exotic — they are the same handful of leaks. A click ID that gets dropped in a redirect, a parameter your tracker expects under one name but the network returns under another, or a confirmation page that reloads and double-counts a single sale. The fix is discipline: trace one real test click end to end and confirm the ID survives every hop, deduplicate on a unique order or transaction ID, and check your logs rather than trusting the dashboard. Running blind without verified tracking is one of the classic beginner mistakes, because it lets you scale a campaign you only think is profitable. If a term here is unfamiliar, the tracking glossary defines each one plainly.
It is a unique string your tracker generates for each click and carries through the funnel. When a conversion is reported, that same ID comes back so the sale can be matched to the exact click that produced it. Networks may label the parameter subid, cid or tid, but it is the same thread.
Usually a broken link in the tracking chain — a click ID lost in a redirect, a mismatched parameter name, or conversions dropped by an ad blocker on a pixel. Trace one test click end to end and check your postback log before blaming the offer.
Only for what S2S cannot give you: retargeting audiences and real-time on-page behaviour. For the conversion count and payout that actually pays you, server-to-server is the reliable source of truth, so most operators run S2S as primary and add a pixel on top.
Once the chain is clean, the next question is which touch earns the credit when a user clicks more than once. That is decided by your attribution model, which can make the same campaign look like a winner or a loser depending on the rule.
If you run campaigns on gut feel, you are flying blind with your own money.
Core · 10 min readWhat a tracking pixel is, how it fires in the browser, how it differs from server-to-server tracking, and where the Meta Pixel, the Conversions API...
Intermediate · 10 min readHow S2S postback tracking works: click ID, macros, where the postback URL goes, testing, deduplication, and why server-side is the durable standard...
