How to read a postback: a practical tracking checklist

Start with one event

A postback is only useful if it maps to something you care about. Before launch, decide which single event the campaign will optimise toward: first deposit, purchase, verified signup, install-to-open. Everything else is a diagnostic, not a target.

The checklist

• Confirm the click ID macro is passed on every landing page and pre-lander.
• Fire a test conversion and verify it appears in the source platform within minutes.
• Check that the postback URL is HTTPS and returns 200, not a redirect.
• Make sure duplicate postbacks are deduplicated by click ID.
• Record the time zone used by both systems. A “yesterday” report that disagrees by a day is usually this.
• Verify that sub-ID values are unique and readable, not hashed beyond recognition.

Reading the report

Do not look at conversion rate alone. Segment by sub-ID and compare three numbers: clicks, conversions and revenue per click. Sources with high clicks and low conversions are either low quality or badly matched to the offer. Sources with decent conversion rates but poor revenue are selling you the wrong customers.

When to blacklist

Give a source enough volume to be judged — usually 300 to 500 clicks — before cutting it. Cutting at 20 clicks produces noise, and noise is expensive. Once a source crosses the threshold and misses your cost-per-event target twice, blacklist it and move on.

The whole system only works if the data arrives clean. That is why tracking deserves an hour of attention before launch, not a five-minute check after.