Share
Embed
BrightonSEO: Lumenlab - Accelerating results with SEO A/B Testing & Edge SEO
Koen Leemans
April 10, 2025
Transcript
56 slides
-
SEO A/B testing isolates the effects of site changes on
organic search performance -
Demonstrate SEO ROI The value of a testing program T12
Months Effect Size) € 4,034,500 Test # Test Type Test Result Organic Clicks Addition Clicks: Addition Clicks: AOV CR rate Revenue Increase* 1 Title Add intent to title tag - PLP Positive 20,000,000 7% 1,400,000 € 40 2% € 1,120,000 2 Title Add category to title tag - PDP Negative 15,000,000 5% 3 Internal links Add block with relevant links - PLP No Change 20,000,000 - 4 Layout Reordering elements - PLP Positive 20,000,000 5% 1,000,000 € 40 2% € 800,000 5 CWV Fixing layout shifts CLS - PLP No Change 20,000,000 - 6 Content Delete content - PLP No Change 20,000,000 - 7 Content Add intro text about category - PLP Negative 20,000,000 1% 8 Schema ProductGroup structured data - PDP Positive 15,000,000 5% 750,000 € 45 3% € 1,012,500 9 Schema ItemList structured data - PLP Positive 20,000,000 4% 800,000 € 40 2% € 640,000 10 Schema Delete negative review snippet - PDP Positive 2,000,000 22% 440,000 € 35 3% € 462,000 Potential Impact on Revenue = Total Traffic to Template * Additional Clicks Percentage) * Average Order Value * Conversion Rate) -
Save time and resources The value of a testing program
T12 Months Effect Size) - € 1,172,500 Test # Test Type Test Result Organic Clicks Addition Clicks: Addition Clicks: AOV CR rate Revenue Increase* 1 Title Add intent to title tag - PLP Positive 20,000,000 2 Title Add category to title tag - PDP Negative 15,000,000 5% 750,000 € 45 3% € 1,012,500 3 Internal links Add block with relevant links - PLP No Change 20,000,000 - 4 Layout Reordering elements - PLP Positive 20,000,000 5 CWV Fixing layout shifts CLS - PLP No Change 20,000,000 - 6 Content Delete content - PLP No Change 20,000,000 - 7 Content Add intro text about category - PLP Negative 20,000,000 1% 200,000 € 40 2% € 160,000 8 Schema ProductGroup structured data - PDP Positive 15,000,000 9 Schema ItemList structured data - PLP Positive 20,000,000 10 Schema Delete negative review snippet - PDP Positive 2,000,000 Potential Impact on Revenue = Total Traffic to Template * Additional Clicks Percentage) * Average Order Value * Conversion Rate) -
Accelerate your learning The value of a testing program T12
Months Effect Size) € 4,034,500 Test # Test Type Test Result Organic Clicks Addition Clicks: Addition Clicks: AOV CR rate Revenue Increase* 1 Title Add intent to title tag - PLP Positive 20,000,000 7% 1,400,000 € 40 2% € 1,120,000 2 Title Add category to title tag - PDP Negative 15,000,000 5% 750,000 € 45 3% € 1,012,500 3 Internal links Add block with relevant links - PLP No Change 20,000,000 - 4 Layout Reordering elements - PLP Positive 20,000,000 5% 1,000,000 € 40 2% € 800,000 5 CWV Fixing layout shifts CLS - PLP No Change 20,000,000 - 6 Content Delete content - PLP No Change 20,000,000 - 7 Content Add intro text about category - PLP Negative 20,000,000 1% 200,000 € 40 2% € 160,000 8 Schema ProductGroup structured data - PDP Positive 15,000,000 5% 750,000 € 45 3% € 1,012,500 9 Schema ItemList structured data - PLP Positive 20,000,000 4% 800,000 € 40 2% € 640,000 10 Schema Delete negative review snippet - PDP Positive 2,000,000 22% 440,000 € 35 3% € 462,000 -
The value of a testing program Every test is a
lesson You discover what drives more traffic. You prevent changes that would have hurt performance. You avoid wasting time and resources. Positive No change Negative -
The value of a testing program Result distribution You discover
what drives more traffic. You prevent changes that would have hurt performance. You avoid wasting time and resources. Positive No change Negative 35% 50% 15% -
Reduce your SEO development tickets by 65%! The value of
a testing program SEO ideas Developer Backlog -
Make you and your organization more efficient by focusing on
what’s proven to work -
How SEO testing works How to launch your own SEO
testing program Control Variant -
How SEO testing works How to launch your own SEO
testing program -
Technical building blocks How to launch your own SEO testing
program Select control and variant pages From experiments to optimizations Apply test changes Analyze results -
A control group provides a baseline to isolate the true
impact of SEO changes. For reliable results, control and variant pages must have similar traffic levels — otherwise, bias can distort outcomes and make attribution unreliable. Why a control group is crucial for reliable SEO tests Select control and variant pages -
Select control and variant pages SEO A/B Page Splitter SEO
A/B Page Splitter A web application for creating two statistically comparable groups for SEO A/B testing. -
Select control and variant pages SEO A/B Page Splitter The
first column should contain all URLs that can be selected as control or variant. The subsequent columns should contain the daily click data for each URL. -
Apply test changes Choose an approach that fits your infrastructure
CMS Existing CRO testing tools Web workers (e.g., Cloudflare) Dedicated SEO A/B testing platforms -
Apply test changes Choose an approach that fits your infrastructure
Dedicated SEO A/B testing platforms -
Even statistically similar groups can behave differently over time. •
Forecasting shows expected performance based on historical data. • The control group adjusts for external factors beyond the test. Why not directly compare variant vs. control? Analyze results -
Analyze results SEO A/B Test Analyzer A tool to analyze
and validate your SEO changes using Causal Impact. -
Analyze results SEO A/B Test Analyzer Organic clicks per day
for the total sum of your control group and variant group pages. For both groups, you need at least 100 days of historical data (data from before the test started), plus all the days the test runs. -
A successful SEO test proves measurable impact — every day
delayed means lost results -
From experiments to optimizations Making changes on the edge Client
CDN Web Worker Origin server Request Request Response Response -
From experiments to optimizations Making changes on the edge HTML
modifications Header modifications <!DOCTYPE html> <html lang="en"> <head> <title>Optimized Edge Title</title> <meta name="description" content="Edge SEO"> <link rel="canonical" href="https://example.com/"> </head> <body> <h1>Welcome</h1> </body> </html> HTTP/1.1 301 Moved Permanently Location: https://example.com/new-page/ Content-Type: text/html Cache-Control: max-age=86400 Content-Length: 162 -
Quickly deploy priority SEO changes while ensuring proven improvements reach
developers — maintaining collaboration and site integrity. Bridge the gap between SEO and development From experiments to optimizations -
Make you and your organization more efficient by focusing on
what’s proven to work — and get results right away -
Request your 30 days of free access! lumenlab.io/brightonseo Want to
talk about what SEO A/B testing can do for you? #58