Case Study: Ad Layout Redesign That Doubled RPM Overnight

Introduction: When Layout Is the Bottleneck

Our team had been running a regional news site for four years, consistently generating around 500,000 monthly pageviews. Despite steady traffic growth, our ad revenue had plateaued at approximately $6,000 per month, yielding an RPM of about $12. We had tried switching ad networks, experimenting with different ad formats, and even increasing our ad density — nothing moved the needle significantly. It was not until we stepped back and fundamentally rethought our ad layout that we achieved a breakthrough: our RPM doubled from $12 to $24 within 48 hours of launching the redesign.

This case study details the exact changes we made, the data that informed our decisions, and the results that followed. If your ad revenue has stagnated despite reasonable traffic, your layout might be the bottleneck holding you back.

The Problem: Analyzing Our Original Layout

Our original layout was typical of many news sites. We had a traditional three-column design on desktop with a header banner, sidebar ads, and a single in-content ad unit. On mobile, the layout collapsed to a single column with ads interspersed between article paragraphs. Here was our original ad configuration:

• Desktop: 728x90 leaderboard in header, 300x250 rectangle in sidebar, one 300x250 in-content ad after paragraph five
• Mobile: 320x50 banner at top, one 300x250 in-content ad after paragraph three
• Total ad units: 3 on desktop, 2 on mobile

On the surface, this seemed reasonable. But when we analyzed our ad performance data in detail, we discovered several critical problems that were suppressing our revenue.

The Data That Revealed the Issues

We conducted a thorough analysis using heatmaps, scroll-depth tracking, and ad viewability reports. The findings were eye-opening:

• Header leaderboard viewability: Only 44%. Users were scrolling past it before it fully rendered, a classic case of banner blindness compounded by slow ad loading.
• Sidebar ad viewability: Just 31%. On our news articles, which typically ran 600–800 words, most users never scrolled far enough for the sidebar ad to become viewable.
• In-content ad viewability: 72%. This was our best-performing unit, but we only had one of them.
• Mobile ad viewability: The top banner achieved only 38% viewability, while the in-content unit hit 68%.
• Average scroll depth: 62% of article length on desktop, 55% on mobile.

The pattern was clear: our best-performing ad position was in-content, where ads appeared naturally within the reading flow. Our worst-performing positions were the traditional placements that users had learned to ignore or that loaded outside the viewport.

The Redesign: Our New Ad Layout

Armed with scroll-depth and viewability data, we designed a new ad layout built around three principles: place ads where users actually look, ensure every ad has high viewability, and match ad density to content length.

Desktop Changes

• Removed the header leaderboard. This was our most controversial decision. The 728x90 banner was generating minimal revenue due to low viewability, and it was pushing our content below the fold. Removing it improved our LCP by 600ms and gave content prime screen real estate.
• Added a sticky sidebar unit. Instead of a static sidebar ad that scrolled out of view, we implemented a 300x600 sticky unit that remains visible as users scroll through articles. This single change increased sidebar viewability from 31% to 89%.
• Increased in-content ads to three. Based on our average article length of 700 words and scroll-depth data, we placed 300x250 ads after paragraphs two, five, and eight. Each ad was spaced far enough apart to avoid feeling overwhelming.
• Added an anchor ad. A slim 728x90 anchor ad fixed to the bottom of the viewport provides persistent visibility without blocking content. This unit alone generates 18% of our total desktop revenue.

Mobile Changes

• Replaced the top banner with a sticky footer ad. The 320x50 top banner had poor viewability and delayed content display. We replaced it with a 320x50 sticky footer unit that achieves 94% viewability.
• Increased in-content ads to two. We added a second in-content unit after paragraph six, giving us two high-viewability placements in the reading flow.
• Implemented a scroll-triggered interstitial. After users scroll past 40% of the article, a non-intrusive interstitial appears between content sections. This unit has a 2.1% click-through rate — our highest across all formats.

Implementation Details

The redesign was implemented over a single weekend to minimize the transition period. We used our ad network's A/B testing tools to validate the new layout before rolling it out site-wide. Here is the implementation timeline:

Friday Evening: Staging Environment

We built the new layout in our staging environment and ran it through Google's ad policy checker to ensure compliance. We also verified that all ad units rendered correctly across Chrome, Firefox, Safari, and Edge on both desktop and mobile devices. Core Web Vitals were tested to ensure the additional ad units did not degrade performance.

Saturday Morning: A/B Test Launch

We launched an A/B test sending 20% of traffic to the new layout. Within four hours, the new layout was outperforming the original by 85% in RPM. We let the test run for 12 hours to ensure statistical significance.

Saturday Evening: Full Rollout

With statistically significant results confirming the improvement, we rolled the new layout out to 100% of traffic. The full rollout showed even stronger results than the A/B test, likely because the ad network's algorithms could optimize more effectively with full traffic volume.

The Results: Before and After

The impact was immediate and dramatic. Here is a detailed comparison of performance metrics before and after the redesign:

Ad Performance Metrics

• Overall viewability: 42% → 76% (improved by 81%)
• Average CPM: $4.20 → $7.80 (improved by 86%)
• RPM: $12.00 → $24.30 (improved by 103%)
• Monthly revenue: $6,000 → $12,150 (improved by 103%)
• Fill rate: 88% → 94% (improved by 7%)

User Experience Metrics

• Bounce rate: 58% → 54% (improved by 7%)
• Pages per session: 2.1 → 2.3 (improved by 10%)
• Average session duration: 2:45 → 3:10 (improved by 15%)
• Core Web Vitals pass rate: 72% → 81% (improved by 13%)

The most remarkable finding was that user experience metrics actually improved despite adding more ad units. By removing poorly performing ads from disruptive positions and placing new ads in natural content flow positions, we created a better reading experience while showing more ads.

Why Viewability Drives Everything

The single most important insight from this redesign is that viewability is the master metric for ad revenue. When advertisers bid on impressions, they increasingly factor viewability into their bid calculations. An ad unit with 90% viewability will attract significantly higher bids than one with 40% viewability, even if they are on the same page. By restructuring our layout around viewability, we did not just increase the number of viewable impressions — we increased the price advertisers were willing to pay for each impression.

This creates a virtuous cycle: higher viewability leads to higher CPMs, which leads to more advertiser demand for your inventory, which further increases competition and CPMs. Our data showed this flywheel effect clearly — CPMs continued to climb for three weeks after the redesign as more demand partners recognized the quality of our inventory.

Mobile-Specific Learnings

The mobile layout changes deserved special attention because mobile traffic represented 65% of our total pageviews. On mobile devices, screen real estate is extremely limited, and every pixel matters. Our old mobile layout with a static top banner was wasting our most valuable viewport space on an ad format with poor viewability. The switch to a sticky footer ad was transformative — it achieved 94% viewability while consuming minimal screen space and never obstructing content.

The scroll-triggered interstitial on mobile was our most debated addition. We were concerned it would frustrate users and increase bounce rates. However, by triggering it only after 40% scroll depth — ensuring users were genuinely engaged with the content — and providing a clear, easy-to-tap close button, we avoided the negative user experience associated with aggressive interstitials. The 2.1% click-through rate confirmed that users encountered the ad at a moment of engagement rather than annoyance. We also configured a frequency cap of one interstitial per user per session to prevent repeated interruptions.

One critical mobile optimization we almost overlooked was ad container sizing. On mobile devices, layout shift is particularly disruptive because content jumps are proportionally larger on smaller screens. We implemented CSS aspect-ratio containers for every mobile ad unit, ensuring that the page layout remained completely stable as ads loaded. This attention to detail improved our mobile CLS score from 0.28 to 0.04, contributing to better mobile search rankings.

Sustaining the Improvement Over Time

Three months after the redesign, our RPM has remained stable at approximately $24, with normal fluctuations based on day of week and seasonal advertiser demand. We attribute this sustainability to the fact that our improvements were structural — better viewability, better placements, better formats — rather than gimmicky short-term tactics. We continue to run small A/B tests each month, testing variations like ad size alternatives and refresh intervals, but the core layout has proven durable. We also established a monthly review meeting where we analyze ad performance data and discuss potential optimizations, ensuring that we never become complacent with our current configuration.

Key Takeaways for Publishers

• Audit viewability before adding ads. More ad units with poor viewability will not increase revenue. Fewer units with high viewability often outperform.
• In-content ads outperform peripheral placements. Ads within the natural reading flow consistently achieve the highest viewability and CPMs.
• Sticky and anchor formats are powerful. These formats maintain visibility throughout the user session, dramatically improving viewability metrics.
• Remove underperforming units. Do not be afraid to cut ad units that are dragging down your overall viewability score. Advertisers evaluate your site holistically.
• Test before you commit. Use A/B testing to validate layout changes before rolling them out site-wide. Data should drive design decisions.

Conclusion

Doubling our RPM overnight was not the result of finding a new ad network or implementing exotic ad technology. It came from a fundamental rethinking of where and how we displayed ads on our pages. By prioritizing viewability over ad count, placing ads in the natural content flow, and removing underperforming units, we transformed our revenue without increasing traffic by a single pageview. For publishers stuck in an RPM rut, a layout redesign guided by viewability data may be the highest-impact change you can make.