← Back
August 17, 2024

Why Google Display Network Has a Placement Quality Problem Google Won't Fix

There's a structural reason the Google Display Network has a placement quality problem, and it has nothing to do with technical limitations. Google's incentive is to serve impressions. Every impression served — on a real news site or a made-for-advertising junk farm — generates revenue for Google. Publishers get paid the same way. The only party in this transaction who needs the impression to actually mean something is you, the advertiser.

This isn't an accusation. It's just the economics. And understanding it explains why the problem hasn't improved despite years of brand safety initiatives, policy updates, and algorithm changes.

What "Low Quality" Actually Means for Your Budget

The term is too abstract. Here's what it means in practice.

A low-quality placement is a site where your ad appeared but almost certainly had no commercial effect. Not because your creative was bad. Not because your targeting was off. But because the site itself exists to collect impressions, not to serve readers.

Made-for-advertising (MFA) sites aggregate content — often AI-generated or scraped from real sources — into a format optimized for maximum ad units per page. High ad density (often above 40%), minimal real content, no editorial voice. A user who lands on one bounces in seconds. Your ad appeared. No one registered it.

AI content farms are a newer variant. They generate plausible-sounding articles at scale and buy traffic through content recommendation networks. The articles are coherent enough to pass automated quality filters. The audience is real but entirely uncommitted — they came from a clickbait headline and leave immediately.

Arbitrage sites buy cheap traffic and expose it to premium-priced ad inventory. The traffic is technically human. The engagement is zero. These are harder to catch because their traffic patterns look cleaner than bot traffic.

Fake news clones mimic the design of legitimate news sites. They attract genuine search traffic on news-adjacent queries. Your ad appears next to content that may be factually incorrect or politically extreme.

The Numbers

Research from HUMAN Security suggests MFA domains account for over 15% of all programmatic display impressions. IAS has reported that nearly 1 in 4 display ads is served to non-human traffic. The ANA estimated that $700M+ per quarter flows to MFA publishers.

On a $10,000/month display budget, conservatively $1,500–$2,500 is going to placements where commercial intent is effectively zero. Not fraud in the legal sense. Just waste.

The problem is worse for campaigns with broad audience targeting or keyword contextual targeting. The wider your net, the higher your exposure to the long tail of the web — which is where MFA sites cluster.

Why Google's Protections Aren't Keeping Up

Google does remove MFA sites from the network. The removal rate can't keep up with the creation rate. AI tools have reduced the cost of launching a convincing-looking content site to near zero. For every domain removed, several more appear.

Content exclusion categories help at the margins. But they're broad and imprecise, and they focus on brand safety — avoiding embarrassment — not on avoiding wasted spend. Those are different problems.

What Actually Works

The placement report in your Google Ads account shows every site your ads ran on. Most advertisers don't look at it regularly. It's long, hard to interpret, and there's no quality score attached to each domain.

The practical move: audit that report monthly for active campaigns. Flag sites with high impression counts but zero conversions. Look for domains you can't verify as legitimate. Check ad density on sites that look suspicious (above 30% is a red flag). Build an exclusion list from your findings and apply it at the account level.

The list accumulates value over time. You won't catch everything — Google will keep serving impressions on sites you haven't excluded yet. But the waste drops measurably, and it's one of the highest-leverage optimizations available on any active Display campaign.

The incentive misalignment isn't going away. Your job is to work around it.