Native Banners

What it is

Native units take their typography, colour, and layout from the publisher page. The shape signals “editorial recommendation,” not “ad.” Lower banner blindness, higher CTR, more variance in the CR distribution.

I didn’t run native at scale at Mobidea — push was the format. So I’ll be honest about the depth: my native data is parallel-buy from a consulting engagement in Q1 2024 (n=1.4M impressions, US + DE + ES, content publisher sites in the AdMaven + Taboola supply mix). The numbers below are from that test, not from a five-year panel.

What the data shows

• CTR median: 0.58–0.94% across the publisher set. Tier-1 display banner CTR for the same offers ran 0.21–0.34%. Native was 1.9–2.7× higher on CTR.
• CR is the catch: CR-given-click on native ran 1.1–1.8% across the same offers. Display CR-given-click ran 2.4–3.6%. Native pulls more clicks; those clicks convert at roughly half the rate of an honest display click.
• End-to-end CR per impression: native at ~0.011%, display at ~0.008% in the same test. Native wins, but the margin is thin — and it widens or collapses fast based on publisher quality.

When native is the right format

• Content publishers — review sites, news, vertical blogs, editorial recommendations
• Offers where the headline can carry curiosity without being misleading (the bridge that holds CR)
• Mid-funnel education-content placements before a hard CTA
• Tier-1 EU and US where editorial publishers have inventory worth buying

When it isn’t

• Brand-safety-strict advertisers — native publisher quality is a long tail; you will see sub-sources you don’t want your logo next to
• Direct-response offers needing a sub-2-second click-to-CR path. Native is mid-consideration.
• Offers where the creative cannot survive “matching the publisher voice” — gambling and adult often can’t

Specs

• Pricing models: CPM · CPC
• Devices: Mobile + Desktop
• Geo coverage: 240+ countries
• Anti-fraud: Multi-layer filtering + ML scoring per impression

For push advertisers reading this: native and push solve different funnel positions. I run them in parallel for the same offer when the LTV math supports both. Don’t compare CPMs in isolation. Compare day-7 CR per dollar.