Google Display Cleaned Up: Retargeting and Custom Segments

Most Google Display prospecting budget in DACH ends up on mobile game apps and made-for-advertising sites. In inherited accounts we routinely find a third to a half of impressions on placements nobody would have bought deliberately. We run display as a retargeting and conquest channel, we clean the placement side before anything else, and if you came here to make display a growth channel we will tell you to put that money into Demand Gen instead. Exclusions first, everything else second: Mobile app inventory is the largest single waste line we find, game apps in particular, where a mis-tap registers as a click and the CTR looks encouraging. We exclude app categories at account level, add parked domains and error pages, and apply a running exclusion list of made-for-advertising domains built from placement reports across the accounts we run. What is left is a smaller and much better-behaved buy than the original media plan assumed, and the first honest cost-per-acquisition figure the account has produced. Reading quality honestly: Click-through rate is not a quality signal on the display network, for the reason above. We read Active View viewable rate and viewable CPM, session depth after the click, and conversion rate by placement group. Frequency is capped deliberately per campaign rather than left at the default, and for retargeting we cap by user and by day. In food and wellness the fastest way to lose a repeat customer is to follow them around for three weeks with the product they already bought. Retargeting split by intent depth: Not one audience. Product page viewed without add-to-cart, cart abandoned before checkout, checkout abandoned at the payment step, and past purchasers approaching a replenishment date are four different situations that deserve four different arguments and four different bids. For consumable wellness and food products, the replenishment window carries the whole case: someone who bought a sixty-day supply should be seen around day forty-five, and not on day three. The one prospecting case we defend: Custom segments built from competitor search terms and competitor URLs, aimed at people demonstrably in-market, running creative that makes an actual comparison rather than a brand statement. It is narrow, it is measurable against a holdout, and we run it with a stop date written into the brief. Creative: Responsive display assets generate combinations your brand team will not enjoy seeing. We upload fixed-size HTML5 or static sets for the placements carrying real volume and keep responsive as fill for the tail. We specify sizes, hierarchy and the German line breaks, because compounds break layouts designed around English word lengths, and we QA every size before launch rather than discovering the problem in a screenshot from your CEO. Consent reality in DACH: German and Austrian publisher inventory sits behind TCF consent management platforms, and addressable reach is meaningfully smaller than the planning tools suggest. We size the buy against consented reach rather than gross reach, so the plan does not promise numbers this market cannot deliver. Not included: Display & Video 360, programmatic buying outside Google Ads, direct publisher deals, and banner production at volume. We write the specification and QA the output; your designer or a freelancer builds. Native advertising platforms are a different buy and we do not run them. Who this is not for: Anyone who wants display to open a new demand channel. Here it protects margin on demand you already created. That is a defensible job and a modest one, and you should hear that before the invoice rather than after. Also not for sites with fewer than a few thousand monthly visitors, where retargeting pools never reach useful size and the whole exercise costs more in management than it returns.