What metrics should I monitor to prevent wasting money on display ads?

Why display waste hides where search waste cannot

Search ads expose the leak through the search-terms report. A bad query has a name and a cost number sitting next to it. Display has no equivalent. The placement report exists, but Google buries it under Insights and Performance Max scrubs most of it into a “Placements where your ads appeared” download with no spend column at all.

The result is predictable. Display and Performance Max display placements absorb between thirty and seventy percent of the budget in most solo-founder accounts, and almost nobody audits them. Six metrics catch the leak before it compounds.

Viewability rate on display impressions

The report: campaigns view with Active View Viewable Impressions divided by total impressions on Display network campaigns, or the Insights tab on Performance Max for display placements.

A viewability rate under sixty percent means more than four in ten impressions never crossed the threshold of being seen by a human for one second. The IAB viewability standard is fifty percent of pixels on screen for one second on display, two seconds on video. Falling below sixty percent on a paid campaign means the inventory is bottom-of-page, behind-the-fold, or auto-refresh slots that an attentive user never registers.

The fix is a placement exclusion list combined with content-category exclusions for low-attention inventory. Below-the-fold and “below 300x250 mobile sticky” exclusions alone often lift viewability ten points inside a week.

Concentration of placement category share

The report: Insights or Where ads showed, sorted by cost or impressions, with mobile-app placements broken out separately.

Healthy display spend concentrates inside the top fifty placements producing more than seventy percent of meaningful engagement. When the top fifty placements account for less than thirty percent of total spend, the campaign is spraying inventory across thousands of sites that each take a few dollars and produce nothing.

A thousand placements taking ten dollars each is the canonical signature of a Performance Max display leak. Pull the placement export, sort by cost descending, and exclude the long tail below a minimum click threshold. The fix takes thirty minutes inside the account-level exclusion list at the MCC.

When audience overlap with retargeting eats prospecting

The report: audience insights at the campaign level, with overlap percentage against Customer Match and remarketing lists.

When a display campaign labeled prospecting shows an audience overlap above thirty percent with the remarketing list, the campaign is buying impressions on people who already converted or already saw a prior ad. The retargeting campaign was already going to reach them at a lower CPM.

This is one of the most common Performance Max problems. The model fires display impressions at warm audiences because they convert faster and make the asset group look efficient, while the prospecting layer the founder thought they were funding gets a fraction of the budget. Segment the spend by audience and the truth surfaces. The Tracking Stack reference covers the de-duplication contract that keeps prospecting and retargeting from double-counting.

Why a high view-through to click ratio inflates ROAS

The report: conversions segmented by attribution type, comparing view-through conversions against click-through conversions on display campaigns.

A view-through conversion fires when a user saw the ad and converted later without clicking. Useful as a directional signal. Dangerous as a primary KPI. A view-through to click ratio above 100 to 1 means the platform is taking credit for almost every conversion in the funnel based on an impression that may have been one pixel tall on a content farm.

Cap view-through conversions at a thirty-day window and exclude them from any ROAS calculation used to set bids. The Google default of ninety days inflates display ROAS in ways that make the leak invisible. A display campaign that looks like four-times return on click-based attribution and twelve-times return on view-through attribution is a campaign that needs the click number used for decisions.

Mobile-app placement spend as a hidden leak

The report: placement type filter set to mobile-app in the Where ads showed view.

Mobile-app placements inside the Google Display Network and Performance Max are the single highest-friction inventory in the system. Children’s games, lock-screen ads, auto-redirect tap traps. Click-through rates look exceptional. Conversion rates collapse. Session duration sits under three seconds because most clicks were accidental.

Mobile-app spend above five percent of total display budget is a leak. The fix is two checkboxes. Exclude mobileappcategory::69500 (Games), exclude the Display Network mobile-app exclusion at the account level. Performance Max requires the same exclusion through the Account-Level Negative Placements list at the MCC, which Google released quietly in 2023 and most founders have never opened.

Session duration from display-source traffic

The report: GA4 traffic-acquisition view, filtered to Google CPC with a campaign filter for display campaigns, looking at engaged sessions and average session duration.

Display-source traffic should clear fifteen seconds of average session duration if the placement quality is real. Below fifteen seconds means the clicks are accidental or the landing page mismatches the ad creative. Below five seconds means accidental clicks dominate, which traces back to mobile-app placements and below-the-fold mobile inventory.

This is the metric that catches what every Google-native report misses. Google Ads measures the click. GA4 measures what happened after. The gap between the two is where display waste compounds.

The metric I look at first

The placement category share. It takes ten minutes to pull and one minute to read. A long-tail of thousands of placements eating ten dollars each is the diagnostic that explains every other metric in this list. Viewability collapses because the long tail is low-quality inventory. Mobile-app spend balloons because Performance Max defaults to including it. Session duration falls because accidental clicks dominate.

Run the free 25-page setup audit for the full diagnostic, use the wasted-spend calculator to size the leak in dollar terms, or read the adjacent diagnostics inside the wasted-ad-spend library. A display layer with three of these six metrics out of band is usually losing twenty to fifty percent of the line item to placements the founder has never seen named.

The Wasted Spend Calculator returns the dollar estimate for the display leak the six metrics describe; the services overview walks the structural fix at the same depth as my paid engagements.