A placement is the exact spot where your ad is shown: a particular website, an app, a widget under an article or a position in a feed. One campaign can run on thousands of placements at once, and they perform wildly differently, so most of the optimisation work is deciding which placements to keep, pay more for, or block.
A placement is the unit of inventory on which an ad is served: a publisher site, a specific widget or zone on that site, a mobile app, a video channel or a position within a social or search feed. Ad platforms expose it under different names (Google Ads “placements” for the Display Network and YouTube, “site ID” or “publisher ID” on native networks, “zone ID” on push and pop networks, “placements” such as Feed, Stories and Reels on Meta) and pass it to trackers as a token in the click URL, so that every click and conversion can be attributed to the inventory that produced it.
Placement-level data is where cold traffic is won or lost. A native or push campaign typically spreads over hundreds or thousands of placements, and results concentrate in a small share of them while a long tail spends money without converting. The standard method is to run broad, wait for enough clicks per placement to read EPC, then build a whitelist (placements that earn, often given higher bids or their own campaign) and a blacklist (placements excluded). Google Ads supports this directly with placement exclusions and exclusion lists shared across campaigns.
Placements also carry brand-safety and fraud risk. The 2017 episode in which major advertisers paused YouTube spending after ads appeared next to extremist content pushed platforms to add placement and content exclusions and pushed buyers to check where their ads actually run. In performance channels the sharper risk is fraud: placements that generate clicks from bots, invisible iframes or click-injection apps. A placement whose click-through rate is high and conversion rate near zero, or whose traffic shares one device profile, is the first candidate for the blacklist.
For affiliates, placements explain why two people running the same offer on the same network report opposite results: they are not buying the same traffic. Passing the placement token to the tracker from day one is the step that makes optimisation possible at all.
A sweepstakes lead offer pays 1.80 USD per approved lead. A push campaign at 0.012 USD per click buys 250,000 clicks (3,000 USD) across 1,900 zones.
| Zone group | Clicks | Leads | Cost | Revenue | Profit |
|---|---|---|---|---|---|
| Top 60 zones (whitelist) | 55,000 | 720 | 660 USD | 1,296 USD | +636 USD |
| Next 340 zones | 95,000 | 610 | 1,140 USD | 1,098 USD | −42 USD |
| Remaining 1,500 zones | 100,000 | 230 | 1,200 USD | 414 USD | −786 USD |
The campaign loses about 190 USD overall while its 60 best zones earn a 96% ROI. Moving those zones into a dedicated campaign with a higher bid, keeping the middle group at a reduced bid and blacklisting the rest turns the next 3,000 USD into a profitable test. None of that is visible without the zone token in the tracking link. Numbers are illustrative.
How to judge traffic quality like an operator: spotting bots and fraud, reading engagement, comparing conversion and approval rate by source, sub-I...
Advanced · 12 min readThe daily and weekly optimisation loop: read the data, decide, then act — rules-based thresholds for scale, hold or kill, funnel-stage priorities,...
Core · 11 min readNative converts — but not if you point cold clicks straight at an offer.
Advanced · 11 min readHow to scale a winning traffic source without breaking it: vertical vs horizontal scaling, budget increments that protect the learning phase, diver...
The specific inventory where an ad appears: a website, a zone or widget on that site, an app, a video channel or a feed position. Platforms report results per placement so you can keep, bid up or exclude each one.
A whitelist is the set of placements you keep or target deliberately because they perform; a blacklist is the set you exclude. Both are built from placement-level EPC and conversion data.
Enough that the expected number of conversions is meaningful. A practical rule is to wait until a placement has spent roughly one to two times the payout without converting before blocking it, more for high-payout offers.
In the traffic source’s reports and as a macro or token you add to your tracking URL, such as site ID on native networks or zone ID on push and pop networks.
References are listed as plain text on purpose; look them up by title and publisher. Updated: 2026-10-06.
Run them inside a network with real tracking, roughly 48-hour payouts and a dedicated manager.
Join the network