LabArticleSeat-measured

Click macros that actually expand

Live-proven on 2026-09-05. 195 optimize-daily landings expanded PLACEMENT_ID, INVENTORY_SOURCE, EXCHANGE_HANDLE, and DOMAIN. /rest/macro cannot enumerate them.

Published 2026-09-09 · Search job: ad click macros placement id

Ad click macros for placement id, inventory source, exchange handle, and domain actually expand on this seat. That is live-proven, not a guess from /rest/macro. On 2026-09-05, 195 optimize-daily landings expanded {{PLACEMENT_ID}}, {{INVENTORY_SOURCE}}, {{EXCHANGE_HANDLE}}, and {{DOMAIN}}, plus the already-known {{CAMPAIGN_ID}}, {{LINE_ITEM_ID}}, and {{CREATIVE_ID}}. /rest/macro still cannot enumerate them.

If you searched for whether a placement-id macro works, use the landing count. Do not use the catalog endpoint.

What we do with an unexpanded value

A value that still contains braces did not expand. Drop it. expandedMacroValue in the Supply Bench logger already does this. Keeping { } in a stored placement or domain key will fail every later join. A naive reconstruct like ${inventory_source}|${placement_id} matches nothing against inventory_agg. The working placement visit key is ${inventory_source}|${exchange_handle}/${placement_id}, for example OPENX|ox/558230702.

If zero reconstructed keys hit an inventory_agg entity in the window, attribution stays none. Do not score every placement as zero lands. That would look like Most placements we bought did nothing without being that pull.

Domain joins lowercase and strip www.. They do not rewrite stored domain entity_key values. Exchange keys are case and whitespace normalized on purpose. A mismatch fails open and leaves flagged supply in the allowlist while telling the buyer it was excluded.

What these macros are not

They are not click tokens. Supply-bench click URLs go straight to /sb/<slug> and do not mint ad_click_tokens or carry an Adprime campaign id. landing_page_visits.campaign_id is null. The Buzz id lives in properties.beeswax_campaign_id. A join on v.campaign_id = c.campaign_id alone reports zero landings against a portfolio that is landing. Clicks, visits, and conversions are different meters.

They are not a fraud certificate. An expanded {{DOMAIN}} of msn.com is a host we can group. It is not an IVT verdict.

What we will not do

We will not publish a guessed macro from /rest/macro 404s or from OpenRTB folklore. We will not treat a brace-containing string as a placement id. We will not add new macros to this list until another dated landing set expands them. The 195 landings are the evidence. A catalog that cannot enumerate macros is not a counter-argument.

How to judge display traffic still holds. Macros give you grain. They do not give you a score. Grain without a visit meter is how a host like msn.com shows up twice before anyone can say whether a person arrived.