+24%
Best performing placement (in-article, after 2nd paragraph)
-31%
Worst performing (sidebar sticky ad)
4/12
Configurations that showed clear positive signals
~500-2,000
Monthly visitors per test site

Hypothesis & Setup

Hypothesis: "If we test different ad placements and formats across SKY ecosystem sites, we will observe clear patterns in CTR and viewability that can inform future monetization strategy — even with low traffic volume."

Setup: 60-day experiment across 3 SKY ecosystem sites (SKY TTS blog, ConverterTools, TrainWithSKY subdomains). 12 different ad configurations tested in 7-day blocks, with 7-day baselines between changes. All sites had 500-2,000 monthly visitors during the experiment period.

Individual Test Results

In-Article (After 2nd Paragraph)

Days 1-7 7 days
+24% CTR vs baseline

Ad placed naturally within content flow. Highest engagement of all tests. Users continued reading after ad.

1.8%
CTR
72%
Viewability

Top of Article (Below Title)

Days 8-14
+18% CTR vs baseline

Strong visibility but users sometimes scrolled past before engaging. Good for brand awareness.

1.6%
CTR
85%
Viewability

Sidebar (Top Position)

Days 15-21
+3% CTR vs baseline

Modest improvement, but highly dependent on page scroll depth. Desktop-only performance.

0.9%
CTR
45%
Viewability

Sticky Sidebar Ad

Days 22-28
-31% CTR vs baseline

Users found sticky ad intrusive on mobile. Increased bounce rate by 12% on affected pages.

0.5%
CTR
68%
Viewability

End of Article (After Content)

Days 29-35
+5% CTR vs baseline

Captured users who finished reading. Low volume but higher intent. Good supplementary placement.

1.2%
CTR
55%
Viewability

Mobile Interstitial (Exit)

Days 36-42
-18% page views per session

High CTR but negative impact on user experience and engagement metrics.

2.1%
CTR
-18%
Engagement

Native In-Feed Ads

Days 43-49
+15% CTR vs baseline

Best balance of monetization and UX. Users perceived as content recommendations.

1.9%
CTR
No negative impact
on bounce rate

Multiple Ad Units (3 per page)

Days 50-56
-22% time on page

Too many ads. Users perceived as "ad-heavy" — decreased session duration significantly.

+35%
Total ad impressions
-22%
Time on page

Other Tests Summary

Days 57-60 + baselines
Mixed / Inconclusive

Responsive vs fixed size: No significant difference (+2%). Auto-ads vs manual: Auto-ads performed worse (-8%). Text-only vs display: Text-only slightly higher CTR (+6%) but lower RPM.

60-Day Experiment Timeline

Week 1
Baseline established — No ads, just user behavior measurement across all sites.
Weeks 2-3
In-content placements tested — After 2nd paragraph performed best (+24% CTR).
Weeks 4-5
Sidebar & sticky tests — Sticky sidebar caused user friction (-31% CTR, increased bounce).
Weeks 6-7
Mobile & native experiments — Native in-feed showed best UX balance.
Week 8
Multi-ad unit test & wrap-up — Too many ads hurt engagement metrics significantly.

Key Takeaways

Placement > Quantity

One well-placed in-content ad outperformed 3-4 ads on the page. After 2nd paragraph was the sweet spot for engagement and monetization.

Mobile Requires Different Approach

Sticky elements and interstitials hurt mobile UX. Native, scroll-friendly placements work best on smaller screens.

Low Traffic = Directional Data Only

With 500-2,000 monthly visitors, we can't claim statistical significance. But patterns emerged clearly enough to guide decisions.

User Experience > Short-Term Revenue

Tests that negatively impacted engagement (sticky ads, interstitials) were abandoned even with higher CTR. Long-term trust matters more.

What Worked & What Didn't

Worked Well

  • In-article ads after 2nd paragraph
  • Native in-feed ads (content recommendations)
  • Top of article placement (below title)
  • Text-only ads on desktop

Didn't Work

  • Sticky sidebar ads (mobile + desktop)
  • Mobile interstitials / exit popups
  • 3+ ad units per page
  • Auto-ads (Google's automatic placement)

Limitations & Caveats

These experiments were conducted on sites with 500-2,000 monthly visitors. Results are directional, not statistically conclusive. Different niches, traffic volumes, and user demographics may yield different results. We're sharing observations, not claiming universal truths.

Back to SKY Labs