Brand Safety & Ad Verification
Protecting Your Investment & Reputation
What is Brand Safety & Ad Verification?
Brand safety is the practice of ensuring that digital ads appear in environments that align with a brand's values and do not harm its reputation. Ad verification is the technology and process of measuring, validating, and ensuring that ads are delivered as intended — to real humans, in viewable positions, in safe contexts, free from fraud.
Together, brand safety and ad verification address three critical questions for advertisers:
- Was the ad seen? (Viewability)
- Was the ad seen by a real human? (Fraud Detection)
- Was the ad seen in a safe environment? (Brand Safety)
Without verification, advertisers risk wasting budget on impressions that are never seen, served to bots, or placed next to harmful content. With the rise of programmatic advertising and the complexity of the supply chain, verification has become essential — not optional — for responsible media buying.
👁️ Viewability: Was the Ad Actually Seen?
Viewability measures whether an ad had the opportunity to be seen by a human user. An impression is counted when the ad is requested and delivered, but that doesn't mean anyone actually saw it. Viewability ensures you're paying for ads that were actually viewable — not hidden below the fold or scrolled past too quickly.
MRC Viewability Standards (Industry Standard)
≥50% of pixels visible for ≥1 continuous second. Applies to banner ads, native ads, and most display formats.
≥50% of pixels visible for ≥2 continuous seconds. Applies to in-stream and out-stream video.
Same standards apply, with additional considerations for mobile viewport, scrolling behavior, and in-app rendering.
Factors Affecting Viewability
- Ad Placement: Above the fold (top of page) vs. below the fold. Above-the-fold ads typically have 70-80% viewability; below the fold can drop to 30-50%.
- Page Load Speed: Slow-loading pages may cause users to scroll before ads render, reducing viewability.
- Ad Position & Size: Larger, well-positioned ads have higher viewability. Sticky or floating ads maintain visibility as users scroll.
- User Behavior: Bounce rate, scroll depth, and time on page all impact viewability.
- Device Type: Mobile viewability often differs from desktop due to smaller screens and different scrolling patterns.
🤖 Ad Fraud & Invalid Traffic (IVT)
Ad fraud refers to deliberate, malicious activity designed to generate fake ad impressions or clicks, stealing advertiser budget. Invalid Traffic (IVT) is the broader category that includes both fraudulent (sophisticated) and non-fraudulent (general) invalid traffic.
Types of Ad Fraud
Automated scripts that simulate human behavior, loading pages and clicking ads. Bots can be simple scrapers or sophisticated headless browsers that mimic real user interactions.
Fraudsters misrepresent inventory as coming from premium publishers when it actually appears on low-quality sites. The buyer thinks they're buying The New York Times but gets a fake site.
Fake clicks on ads, often generated by bots or click farms, designed to drain budgets on CPC campaigns. Common in search and performance advertising.
Ads loaded in 1x1 invisible pixels or stacked behind other content. The impression is counted but never seen. Often used in display and video fraud.
Malware that inserts unauthorized ads into websites, replacing legitimate inventory. The publisher and advertiser both lose.
Hundreds or thousands of real mobile devices stacked in racks, running automated scripts to generate fake impressions and clicks. Harder to detect than bot traffic.
General vs. Sophisticated Invalid Traffic (GIVT vs. SIVT)
| Type | Description | Examples |
|---|---|---|
| GIVT | General Invalid Traffic. Easier to detect and filter using basic lists and heuristics. | Crawlers, known bots, data center traffic, pre-fetch requests |
| SIVT | Sophisticated Invalid Traffic. Advanced fraud that requires machine learning and behavioral analysis to detect. | Headless browsers, device farms, click farms, domain spoofing, pixel stuffing |
- Behavioral Analysis: Identifying patterns inconsistent with human behavior (too fast, too regular, too perfect)
- Honeypots: Decoy sites and ad placements designed to attract bots
- IP Blacklists: Blocking known data centers and suspicious IP ranges
- Device Fingerprinting: Identifying anomalous device characteristics
- Machine Learning: Models trained on billions of data points to identify fraud patterns
🛡️ Brand Safety & Suitability
Brand safety ensures ads do not appear alongside content that could harm a brand's reputation. This includes explicit categories like violence, hate speech, adult content, and misinformation, as well as adjacent concerns like political content or sensitive news topics.
Brand Safety Categories (IAB & GARM Standards)
Violence, hate speech, terrorism, harassment, dangerous acts, self-harm
Pornography, nudity, sexual content, dating sites
Drugs, weapons, counterfeit goods, piracy
Natural disasters, crime, terrorism, war, political controversy
Discrimination based on race, religion, gender, orientation, ethnicity
Fake news, conspiracy theories, misleading content
Brand Safety Controls
- Blocklists: Pre-defined lists of domains, apps, or URLs where ads should never appear. Can be custom (advertiser-specific) or curated by verification vendors.
- Allowlists: Approved list of domains or apps where ads may appear. Provides maximum control but limits scale.
- Category Exclusions: Block entire content categories (e.g., "violence," "adult," "hate speech") using IAB categories or custom taxonomies.
- Keyword Blocking: Block pages containing specific keywords (e.g., "shooting," "bomb," "terror"). Can be applied at page-level in real-time.
- Contextual Targeting: Serve ads only on pages that match specific contextual criteria (e.g., "sports," "cooking," "travel") while automatically avoiding unsafe categories.
- Suitability Controls: Instead of binary safe/unsafe, some platforms offer "suitability" tiers — allowing brands to choose their comfort level with sensitive content (e.g., avoid "breaking news" but accept general news).
🔍 Major Ad Verification Vendors
Verification vendors provide independent measurement of viewability, fraud, and brand safety. They work across DSPs, SSPs, and ad servers, providing unified reporting and pre-bid blocking capabilities.
Full-suite verification including viewability, fraud detection, brand safety, and contextual analysis. Strong in pre-bid blocking and supply chain transparency. MRC-accredited across all core metrics.
Leading verification platform with strong fraud detection and brand safety solutions. Known for pre-bid avoidance and Authentic Impression metric. MRC-accredited.
Strong in viewability measurement, attention metrics, and fraud detection. Integrated with major DSPs for pre-bid filtering. Known for Moat Analytics dashboard.
Specialized in sophisticated fraud detection. Focuses on SIVT, bot detection, and account fraud. Used by major platforms and exchanges for fraud prevention.
Focus on fraud analytics, supply chain transparency, and compliance. Known for ad fraud reports and seller trust indices.
Specialized in malware detection, ad quality, and security. Focuses on blocking malicious creatives and protecting user experience.
Pre-Bid vs. Post-Bid Verification
| Approach | Description | Pros | Cons |
|---|---|---|---|
| Pre-Bid | Verify inventory before bidding. DSP checks against verification vendor's data to block unsafe or non-viewable impressions before the bid is submitted. | Prevents wasted spend, protects budget, improves efficiency | May reduce scale if filters are too restrictive |
| Post-Bid | Verify after impressions are served. Measure viewability, fraud, and brand safety after the ad runs. | Comprehensive measurement, can optimize future campaigns | Doesn't prevent bad impressions, only measures them |
Best Practice: Use both. Pre-bid for fraud and brand safety protection; post-bid for measurement, reporting, and optimization.
📋 Industry Standards & Certifications
Several organizations set the standards for verification, viewability, and fraud detection. Advertisers should require MRC-accredited measurement and look for TAG certification.
Media Rating Council (MRC)
The MRC is the industry's auditing and accreditation body. They establish measurement standards and accredit vendors that meet those standards. Advertisers should prioritize MRC-accredited measurement for viewability, fraud, and brand safety.
- Viewability Accreditation: Vendors must prove their measurement matches MRC standards
- Fraud Detection Accreditation: Rigorous testing of detection methods
- Invalid Traffic Accreditation: Both GIVT and SIVT detection capabilities
Trustworthy Accountability Group (TAG)
TAG is the industry's anti-fraud certification program. TAG-certified companies have been vetted for fraud prevention practices.
- TAG Certified Against Fraud: Companies that meet anti-fraud standards
- TAG Certified Against Malware: Companies with malware prevention
- TAG Data Seal: Companies with data protection practices
Global Alliance for Responsible Media (GARM)
GARM is a cross-industry initiative that established common definitions and frameworks for brand safety and suitability. Their "Brand Safety Floor" defines categories that should be universally avoided, while "Suitability" allows for nuanced control.
- ☐ Use MRC-accredited verification vendors
- ☐ Implement pre-bid fraud and brand safety blocking
- ☐ Set viewability benchmarks (e.g., 70% for display, 60% for video)
- ☐ Define brand safety categories based on GARM framework
- ☐ Review supply chain transparency (ads.txt, sellers.json)
- ☐ Monitor fraud rates and invalid traffic by publisher
- ☐ Regularly audit verification reports and optimize accordingly
📰 Publisher Best Practices for Brand Safety
Publishers also play a critical role in brand safety. Unsafe inventory drives away premium advertisers and reduces CPMs. Publishers who prioritize brand safety attract higher-quality demand and command higher rates.
Publisher Responsibilities
- Implement ads.txt & sellers.json: Transparency in authorized sellers prevents domain spoofing and builds trust with buyers.
- Content Moderation: Monitor user-generated content (comments, forums) to prevent toxic content next to ads.
- Ad Quality Controls: Block malicious creatives, malware, and inappropriate ads from appearing on your site.
- Inventory Categorization: Properly classify content using IAB categories to help advertisers make informed decisions.
- Work with Verification Vendors: Allow verification tags to run on your inventory. Blocking verification is a red flag for buyers.
- Disclose Made-for-Advertising (MFA) Sites: MFA sites (designed primarily for ad arbitrage) are increasingly avoided by premium advertisers.
⚙️ How Verification Works in Programmatic
Verification is integrated throughout the programmatic supply chain, from pre-bid blocking to post-bid measurement.
1. Pre-Bid (DSP Level): DSP calls verification vendor API with bid request data (domain, app, content categories). Vendor returns score or block decision. DSP can choose to bid or skip based on response.
2. Auction & Win: If bid wins, impression is served with verification tag embedded.
3. Post-Bid (Ad Served): Verification tag executes in user's browser, measuring viewability, scanning page content, detecting fraud signals.
4. Reporting: Data sent back to verification vendor, aggregated, and made available to advertiser through dashboard and API.
5. Optimization: Advertisers use verification data to optimize campaigns — adjusting bids, blocking bad domains, refining targeting.
Verification Macros & Tags
DSPs and ad servers support macros that pass impression-level data to verification vendors. Common macros include:
- ${AUCTION_PRICE}: Clearing price of the impression
- ${PUBLISHER_DOMAIN}: Domain where ad appears
- ${AD_ID}: Unique creative identifier
- ${CAMPAIGN_ID}: Campaign identifier
📚 Quick Reference: Verification Metrics & Benchmarks
≥50% visible for ≥1 second
Benchmark: 65-75%
≥50% visible for ≥2 seconds
Benchmark: 60-70%
Fraudulent + non-fraudulent invalid
Benchmark: 5-15% (varies by channel)
Ads appearing on unsafe content
Benchmark: <1-3% (with pre-bid)
Gold standard for measurement
Requirement: Required for trusted reporting
Anti-fraud certification
Requirement: Recommended for partners