Paid traffic sources through one connection.

Choose sources by campaign objective, audience availability, format, market and measurement needs. A source is suitable when its delivery can be separated and evaluated against the same accepted event.

Source-level detail helps an advertiser identify placements that support or weaken the campaign result. It also makes exclusions and follow-up tests more precise than one blended average.

Compare formats under consistent audience and destination conditions, then judge them through the same business event. Impression and click rates have different meanings and should not be ranked as if they were interchangeable.

Confirm that the page matches the ad, works on the intended devices and records the chosen event. Source parameters should survive redirects so the test remains attributable.

Divide the budget into named source or placement cells with an agreed review point. Avoid spreading a small test across so many variables that no cell can support a decision.

Capture campaign, source, placement, creative and device identifiers alongside the approved outcome. Use stable naming so reports from the ad platform and analytics system can be reconciled.

Evaluate readable sessions, destination behavior, accepted events and rejection signals by source. Media cost alone cannot show if the visitor fits the offer or completes a useful action.

Preserve request timing, source IDs, device patterns and downstream event records before escalating a concern. First rule out tagging errors, broken pages and measurement duplication.

Stop the placement when its traceable cost, rejection pattern or downstream quality crosses the written limit. Record the source ID and evidence, then keep it excluded until a controlled review supports another test.

Add budget after accepted results repeat and source mix remains visible at the higher delivery level. Increase one control gradually, then confirm that cost and quality stay within the approved boundary.