What is a tracker?

What a tracker does

A tracker sits between the ad and the destination. The user clicks the ad, arrives at the tracker URL, the tracker records the interaction — campaign, source, creative, timestamp, device context — and then redirects to the real destination.

That redirect step is the whole mechanism. It is also the whole fragility: every hop is a place where something can be lost, and the user is waiting through all of it.

What a tracker records

• Campaign structure. Network, campaign, ad group, creative — whatever taxonomy you encode.
• Timestamp, which starts the attribution window.
• Device and context available at click time.
• The destination, including any deep link to a specific in-app screen.

How trackers break

Redirect chains. Each additional hop adds latency and another chance of failure. A user on a poor connection abandons before arriving, and the click is counted while the visit is not.

Parameter stripping. Some browsers and platforms remove tracking parameters in transit, which takes the attribution with it.

Copy-paste drift. Trackers are long and hand-edited, so a mistyped campaign parameter lands traffic in the wrong bucket. A generator and a naming convention prevent more measurement problems than any analytics tool fixes.

Broken deep links. The redirect resolves to a web page rather than the intended app screen, and the experience the campaign promised is not delivered.

Where trackers stop working

A tracker requires a click. On AI surfaces a substantial share of influence produces no click at all — the user reads a recommendation and acts later, or an agent completes the purchase without a browser ever being involved.

For those paths there is no URL for a tracker to sit in front of, and no amount of tracker hygiene recovers the signal. Measurement moves server-side and, for the causal question, to incrementality.