Eight Signals That Predict Whether a Display Placement Is Wasting Your Budget
There are over two million websites in the Google Display Network. Your campaigns will touch thousands of them. Some will deliver real results. Others will absorb budget silently. The challenge is that the difference isn't always obvious from a domain name or a quick visit.
Practitioners and researchers studying display quality have converged on a set of signals that reliably predict whether a placement will produce commercial results. Here they are.
1. Ad Density
The number of active ad units per page is one of the strongest predictors of placement quality. MFA sites maximize ad units — they'll run 8 to 15 ads per page. IAS research found that MFA sites average 11.4 ads per page compared to 4.1 on legitimate publisher sites.
High ad density creates two problems: your ad competes for attention against a dozen others on the same page, and the ad-saturated environment signals to users that the site isn't a trusted destination. Both reduce effectiveness.
Flag at: More than 5–6 ad units visible on a standard page layout.
2. Content-to-Ad Ratio
Related to density but distinct. Some sites have a low raw ad count but very thin content. A 200-word article with three ad units has a different ratio than a 2,000-word piece with three ad units. Sites with minimal content and significant ad presence are usually built for inventory, not readership.
3. Traffic Source Composition
Where does a site's traffic come from? Sites with strong direct and organic search traffic have audiences with real intent. Sites that buy 50%+ of their traffic through content recommendation networks (Outbrain, Taboola, MGID) are arbitraging: they pay for visitors and monetize them through ad impressions.
This model produces real human traffic with near-zero commercial intent. Users didn't seek out the site. They clicked a headline in a content feed, glanced at the page, and left.
4. Post-Click Bounce Rate and Session Duration
Post-click behavior on your landing page tells you whether the audience was genuinely engaged. Placements where users consistently bounce within 5–10 seconds have audiences that weren't in any purchasing mindset.
Compare each placement against your campaign average. Placements with bounce rates 30+ percentage points above average, sustained over 60+ days, are strong exclusion candidates regardless of other signals.
5. Conversion Rate on a 90-Day Window
The most direct signal. If a placement has received meaningful spend over 90 days and produced no conversions, something is wrong. Combined with high bounce rate, it's a clear exclusion signal.
Don't cut placements too early. Some need volume to show statistical patterns. Ninety days and a minimum spend threshold (roughly $100–200 depending on your CPA target) is a reasonable starting point.
6. Internal Link Depth
Real publications have editorial depth. They have category archives, author pages, related articles, real internal search. MFA sites are shallow — navigation links often lead back to the homepage or to more thin-content pages.
Test this manually in 60 seconds: click several internal links. Does the content get richer as you go deeper, or does every page look like the same template with different keywords?
7. Author and Editorial Credibility
Legitimate publishers have identifiable human editors and writers. Author pages link to people with real professional histories. MFA sites use generic bylines ("Staff Writer," "Admin") or AI-generated personas. Byline names that don't appear anywhere outside the site are a red flag.
8. Social Proof
Real audiences share content. A publisher with a genuinely engaged readership will have social presence with follower counts and engagement rates proportional to their claimed traffic. A site claiming 5 million monthly readers with 400 Twitter followers and no post engagement doesn't have the audience it appears to.
Low-quality sites often have social icons that link to abandoned accounts or pages with minimal engagement despite high claimed traffic.
No single signal is definitive. Placements that fail three or more of these checks consistently are spending your budget without delivering commercial results.
Running through this checklist on your highest-spend unknown placements — even manually — will identify the worst offenders faster than waiting for conversion data to accumulate. Start with the top 20 placements by spend that you can't immediately verify as legitimate publishers. The list of domains worth excluding becomes clear quickly.