Native advertising research is different from other traffic types because the creative format has two distinct components that must work together: the thumbnail image and the headline. Both need to perform in the "recommended content" widget context - they need to blend with editorial content while still driving clicks from high-intent readers.
This walkthrough is specific to native ad research using AdPlexity Native. It covers the full workflow from filter setup through creative pattern analysis to publisher site research.
Setting Up for Native Research
Start with the baseline filter configuration from the Find Winning Campaigns tutorial, then add native-specific refinements:
| Filter | Native-Specific Setting | Reason |
|---|---|---|
| Ad Network | Taboola or Outbrain (separately) | Taboola and Outbrain have different audience demographics - research them separately |
| Country | United States first, then expand | US has highest CPC and most developed native market - benchmark against US data first |
| Device | Desktop only first | Native creative on desktop vs mobile can differ significantly - don't mix |
| Run Duration (min) | 30 | Standard profitability signal |
| Last Seen | Past 7 days | Confirm still active |
Headline Analysis
Native headlines have a specific psychological structure. They appear as "article recommendations," so they must look like editorial content while still driving clicks. After analyzing hundreds of high-performing native headlines, these are the consistent patterns:
Curiosity Gap Headlines
The most common pattern. Creates an information gap the reader needs to close:
- "The One Thing Cardiologists Never Eat (And Why)"
- "What Happens to Your Body When You Stop Eating Bread"
- "Why Millionaires Keep Their Money in This Account"
Structure: [Authority + topic] + [Implied secret or contrast]
Specific Number Headlines
Numbers create specificity and credibility:
- "Man Loses 47 Pounds With This 3-Step Protocol"
- "Woman Pays Off $31,000 in Debt Using This Method"
- "Retirees Are Earning $2,340/Month Doing This From Home"
Structure: [Specific person] + [Specific number result] + [Implied method]
Newsjacking Headlines
Tie to current events or trending topics:
- "What [Celebrity] Uses to Look 20 Years Younger"
- "IRS Rule Change Helps Seniors Keep More Money"
Look for the number-in-headline pattern in long-running campaigns. Specific numbers ("47 pounds," "$2,340") test dramatically better than round numbers ("50 pounds," "$2,500") because they feel more credible and specific. If you see this pattern in 60-day+ campaigns, that specificity is likely winning CRO test data.
Image Pattern Analysis
Native ad images must work in a 200x200px to 400x250px thumbnail format. The image and headline must create combined curiosity. Common winning image patterns by vertical:
Health/Nutra
- Before/after splits: Clear visual transformation, usually featuring real people (not stock models)
- Close-up focus: Extreme close-up of the "problem" (stomach, skin, joint) - creates visual specificity
- Natural imagery: Herbs, berries, supplements in natural settings - triggers "natural solution" association
- Surprising visuals: Image that creates visual dissonance with the headline (e.g., headline about belly fat + image of a doctor or scientist)
Finance
- Regular people: Not models, not office-stock-photo people. Real-looking retirees, middle-aged couples
- Paper documents: Checks, envelopes, bank statements create credibility for financial claims
- Location signals: Geographic imagery ("people in [state]...") or flag imagery for country-specific offers
Gaming/Entertainment
- Game screenshots: Real in-game images with exaggerated elements
- Comparison shots: "New vs Old" or "Free vs Premium" side-by-side
Use AdPlexity's grid view (not list view) when analyzing image patterns. Set a high result count and scan thumbnails visually without clicking. After viewing 50+ ads in a niche, you'll recognize the dominant image patterns immediately. The most-used patterns are usually the proven ones.
Publisher Site Analysis
For each high-performing campaign, click through to see the publisher list - which sites the ad ran on. This is underused research data.
What publisher data tells you:
- Editorial fit: Is the ad running on news sites, entertainment sites, or niche verticals? The editorial context affects audience intent and CTR
- Tier of publisher: Premium placements (CNN, CNBC, Forbes) vs long-tail placements - premium costs more but delivers better-quality clicks in some verticals
- Publisher whitelist strategy: If a campaign runs on 200 sites, it's using broad targeting. If it runs on only 20 specific premium sites, the advertiser has built a whitelist of their best performers
Taboola vs Outbrain Research
These are the two largest native networks and they have meaningfully different audience demographics:
| Characteristic | Taboola | Outbrain |
|---|---|---|
| Publisher quality | Mix (high to lower tier) | Skews premium publishers |
| CPCs | Generally lower | Generally higher |
| User demographics | Broader, mass market | Slightly more affluent/educated |
| Best verticals | Health, gaming, broad finance | Finance, insurance, B2B |
| Research approach | Filter Taboola-only | Filter Outbrain-only |
Research each separately in AdPlexity. A campaign structure that works on Taboola may need different creative on Outbrain.
MGID Research (Tier-2/3 Focus)
MGID is the leading native network for Tier-2 and Tier-3 countries. If your campaigns target Southeast Asia, Eastern Europe, LATAM, or MENA, MGID-specific research in AdPlexity is essential.
Key differences in MGID research:
- CPCs are dramatically lower (often $0.01-0.05 vs $0.30-1.00 on Taboola US)
- Volume is high but conversion rates and payouts differ by country
- Local language creatives often outperform English even in English-capable markets
- Sweepstakes and pin-submit offers dominate MGID Tier-3
Start Your Native Research
Pick one vertical, one network, one country. Do the complete research workflow. Your first session builds the baseline pattern library you'll reference for months.
Get AdPlexity Native →