SEO Data Analysis: Turn Data Into Strategy

What You'll Master

• Advanced GA4 event tracking for SEO
• Custom SEO dashboards in Looker Studio
• Statistical significance testing
• Traffic forecasting and modeling
• Cohort analysis for organic users
• Attribution modeling

Scroll Depth Tracking

Why it matters: Shows if users actually read your content or bounce after seeing the title. High rankings mean nothing if users don't engage.

How to set up:

1. In GA4, go to Admin → Data Streams → Enhanced Measurement
2. Enable "Scrolls" (tracks 90% scroll by default)
3. For custom thresholds (25%, 50%, 75%), add via Google Tag Manager:

   window.addEventListener('scroll', function() {
     var scrollPercent = (window.scrollY / (document.body.scrollHeight - window.innerHeight)) * 100;
     if (scrollPercent >= 25 && !window.scroll25) {
       window.scroll25 = true;
       gtag('event', 'scroll_25', { page_path: window.location.pathname });
     }
     // Repeat for 50, 75, 100
   });

4. Create custom event parameters to segment by page type (blog, product, landing page)

Analysis: Pages with high rankings but low scroll depth need content improvements. If only 20% of users scroll to 50%, your intro isn't compelling enough.

Internal Search Tracking

Why it matters: What users search for on your site reveals content gaps and keyword opportunities you're missing.

How to set up:

1. Go to GA4 → Admin → Data Streams → Configure tag settings → Show all → Define internal site search
2. Add your search query parameter (usually ?s= or ?q=)
3. Create a custom dimension: Admin → Custom Definitions → Create custom dimension
   • Dimension name: "Search Term"
   • Scope: Event
   • Event parameter: search_term

Analysis: If users frequently search for "keyword difficulty calculator" but you don't have that page, create it. High search volume + zero results = content opportunity.

CTA Click Tracking

Why it matters: Measures conversion intent from organic traffic. Not all traffic is equal - track which pages drive action.

How to set up in GTM:

1. Create a trigger: Click - All Elements → Some Clicks → Click Classes contains "btn-primary" (or your CTA class)
2. Create a GA4 Event tag:
   • Event name: cta_click
   • Event parameters:
     • cta_text: {{Click Text}}
     • cta_location: {{Page Path}}
     • traffic_source: organic (if medium = organic)

Analysis: Calculate CTA click rate by landing page. If your product comparison page gets 10,000 organic visits but only 50 CTA clicks (0.5%), the content isn't converting - improve your pitch or CTA placement.

Essential Custom Dimensions

• Content Category: Tag pages as "blog," "product," "tutorial," "comparison" to analyze performance by content type
• Word Count: Pass article length as a dimension to correlate content depth with rankings/engagement
• Author: Track which authors' content performs best (engagement, conversions, rankings)
• Keyword Theme: Group pages by keyword clusters (e.g., "keyword research," "link building") to measure topical authority impact
• Publish Date: Analyze content decay - do pages lose traffic over time? How quickly?
• Target Keyword: Tag each page's primary keyword to track individual keyword performance beyond Search Console's 1000-row limit

Step 1: Connect Data Sources

1. Go to Looker Studio (lookerstudio.google.com)
2. Create a new report → Add data → Google Analytics (GA4)
3. Add another data source → Search Console
4. Optional: Add Semrush, Ahrefs API via Google Sheets connector for ranking data

Step 2: Organic Traffic Overview Widget

What to include:

• Scorecard: Total organic sessions (compared to previous period)
• Time series chart: Daily organic sessions with trend line
• Filter: Add date range comparison (This period vs. Previous period)
• Segment: Break down by New Users vs. Returning Users

Why it matters: Identifies traffic trends at a glance. Sudden drops = algorithm updates or technical issues. Gradual decline = content decay or competitor gains.

Step 3: Top Landing Pages by Traffic & Conversion

Metrics to display:

Column	Metric	Why It Matters
Landing Page	Page Path	Shows which pages drive traffic
Sessions	Organic sessions	Traffic volume
Avg. Engagement Time	Avg. engagement time	Content quality indicator
Conversions	Key events (formerly conversions)	Business impact
Conversion Rate	Calculated: (Conversions / Sessions) * 100	Traffic quality

How to use: Sort by sessions descending to find your top performers. Then sort by conversion rate to find high-intent pages. Low traffic + high conversion rate = opportunity to build more content around that keyword theme.

Step 4: Search Console Performance Widget

Data from Search Console connector:

• Total clicks from organic search
• Average position across all queries
• CTR (Click-through rate)
• Impressions (how often you appear in search)

Create a table showing:

• Query (keyword)
• Clicks
• Impressions
• CTR
• Average Position

Analysis insights:

• High impressions + low clicks + position 1-3 = poor title/meta description, needs optimization
• Position 4-10 + high impressions = optimize to break into top 3 for massive traffic gain
• Position 11-20 = "striking distance" keywords - small improvements can push to page 1

Dashboard Template

Save time with this layout:

1. Row 1: Scorecards (Organic Sessions, Conversions, Avg. Position, Total Clicks)
2. Row 2: Traffic trend chart (GA4) + CTR trend chart (GSC)
3. Row 3: Top Landing Pages by Sessions table
4. Row 4: Top Queries by Clicks table (from Search Console)
5. Row 5: Content Type breakdown (blog vs. product vs. tutorial performance)

Common SEO Scenarios Requiring Statistical Tests

• Testing if a meta description change increased CTR
• Measuring if adding schema markup improved rankings
• Determining if internal linking boosted page authority
• Verifying if page speed improvements reduced bounce rate
• Checking if content updates increased engagement time

Example: Did Your Title Change Improve CTR?

Scenario: You changed a product page title. Before: 5,000 impressions, 250 clicks (5% CTR). After: 5,000 impressions, 300 clicks (6% CTR).

Question: Is the 1% improvement statistically significant or random chance?

Use a Chi-Square calculator:

1. Go to a free online chi-square calculator (socscistatistics.com, for example)
2. Input your 2x2 table:

                 Clicked    Didn't Click    Total
   Before:          250         4,750        5,000
   After:           300         4,700        5,000

3. Check the p-value. If p < 0.05, the change is statistically significant (95% confidence)

Result interpretation: In this case, p ≈ 0.08, which means the improvement is NOT statistically significant. The 1% CTR increase could easily be random variance. Don't celebrate yet - test longer or try a more dramatic title change.

Example: Did Content Expansion Increase Engagement?

Scenario: You expanded a blog post from 800 to 2,000 words.

• Before (30 days): Average engagement time = 45 seconds, SD = 20 seconds, n = 500 users
• After (30 days): Average engagement time = 62 seconds, SD = 25 seconds, n = 520 users

Use a T-Test calculator:

1. Go to a free t-test calculator (graphpad.com/quickcalcs/ttest1.cfm)
2. Enter: Mean 1 = 45, SD 1 = 20, n1 = 500 / Mean 2 = 62, SD 2 = 25, n2 = 520
3. Check p-value

Result: p < 0.001 - highly significant! The longer content genuinely increased engagement. This validates your content strategy.

Statistical Significance Rules for SEO

• p < 0.05: Statistically significant (95% confidence the change is real)
• p < 0.01: Highly significant (99% confidence)
• Sample size: Need at least 100 conversions or 1,000 sessions for reliable tests
• Test duration: Run tests for at least 2-4 weeks to account for weekly seasonality
• Don't p-hack: Don't keep testing until you get p < 0.05. Decide test duration upfront.

Method 1: Linear Regression (For Steady Growth)

When to use: Your traffic has been growing steadily month-over-month without major spikes or drops.

How it works: Finds the line of best fit through your historical data and projects it forward.

Excel/Google Sheets method:

1. Export 12 months of organic traffic data from GA4 (Sessions by Month)
2. Create a column for month numbers (1, 2, 3, ..., 12)
3. Use formula: =FORECAST.LINEAR(13, B2:B13, A2:A13)
   • 13 = next month
   • B2:B13 = your traffic values
   • A2:A13 = month numbers
4. This predicts month 13's traffic based on linear trend

Example: If you grew from 10,000 → 15,000 sessions over 12 months (linear growth), forecast predicts ~15,400 for month 13.

Method 2: Exponential Growth (For Compounding Content)

When to use: Your traffic accelerates as you build topical authority and internal linking creates compounding returns.

Formula: Future Traffic = Current Traffic × (1 + growth rate)^months

Example:

• Current monthly traffic: 20,000 sessions
• Average month-over-month growth rate: 8%
• Forecast for 6 months: 20,000 × (1.08)^6 = 31,700 sessions

In Excel: =20000 * (1.08)^6

Method 3: Keyword-Level Forecasting

Most accurate method: Bottom-up forecasting based on individual keyword opportunities.

Process:

1. List all keywords you're targeting (from Semrush, Ahrefs, GSC)
2. For each keyword, note:
   • Current position
   • Search volume
   • Expected position in 3-6 months (based on difficulty and your resources)
   • Expected CTR at that position (use CTR curves: position 1 ≈ 30%, position 3 ≈ 15%, position 5 ≈ 8%)
3. Calculate predicted traffic: Search Volume × Expected CTR
4. Sum across all keywords

Example calculation:

Keyword: "best keyword research tools"
Search volume: 5,000/month
Current position: #8 (CTR ≈ 3% = 150 clicks/month)
Target position: #3 (CTR ≈ 15% = 750 clicks/month)
Expected traffic gain: +600 clicks/month

Seasonal Adjustments

Don't forget to account for seasonality. Keyword research tools might spike in January (New Year resolutions) and drop in summer.

How to adjust:

1. Calculate seasonal index for each month: (Month Traffic / Average Monthly Traffic)
2. Apply to forecast: Forecasted Traffic × Seasonal Index
3. Example: If June typically sees 85% of average traffic, multiply June forecast by 0.85

Setting Up Cohorts in GA4

1. Go to GA4 → Explore → Cohort Exploration
2. Set cohort by: "First User Source/Medium"
3. Filter to: source/medium contains "google / organic"
4. Choose cohort size: Week or Month
5. Select metric: Active Users, Engaged Sessions, or Key Events (conversions)
6. Set time granularity: Daily or Weekly

How to Read Cohort Tables

Example cohort table:

Cohort        Week 0    Week 1    Week 2    Week 3    Week 4
Jan Week 1    1,000     120       80        60        50
Jan Week 2    1,200     140       95        70        55
Jan Week 3    1,100     110       75        55        45

What this tells you:

• Retention rate Week 1: 120/1,000 = 12% of users return within a week
• By Week 4, only 50/1,000 = 5% are still active
• If retention is improving over time (Jan Week 3 cohort has higher retention than Jan Week 1), your content quality is improving

Cohort Analysis Insights for SEO

• Low Week 1 retention (< 5%): Your content doesn't build loyalty. Users search → land → leave forever. Solution: Add email signup, related content, remarketing.
• High Week 1, low Week 4: Good initial value, but users forget about you. Solution: Build brand awareness, create content series, encourage bookmarking.
• Improving retention over time: Your SEO is attracting better-fit users. Content quality or targeting has improved.
• Declining retention: Recent traffic is lower quality. Check if you're ranking for wrong keywords or if content quality dropped.

Common Attribution Models

Model	How It Works	When to Use
Last Click	100% credit to final touchpoint before conversion	Simple, but undervalues SEO (users often discover via organic, convert via direct)
First Click	100% credit to first touchpoint	Gives SEO full credit for discovery, but ignores nurturing channels
Linear	Equal credit to all touchpoints	Fair middle ground, good starting point for multi-channel analysis
Time Decay	More credit to recent touchpoints	Best for short sales cycles (e-commerce, SaaS trials)
Data-Driven	ML algorithm assigns credit based on actual impact	Most accurate, but requires significant conversion volume (1000+/month)

Setting Up Attribution in GA4

1. Go to GA4 → Advertising → Attribution → Attribution Settings
2. Choose your lookback window (default 30 days for most businesses)
3. Select attribution model:
   • Data-driven (if you have enough conversions)
   • Linear (if you don't)
4. Go to Advertising → Model Comparison to see how different models value SEO

Example: How Attribution Changes SEO's Value

Scenario: User journey

1. Day 1: Discovers your site via Google organic search (reads blog post)
2. Day 5: Returns via email newsletter (you captured their email)
3. Day 12: Comes back via direct/bookmark, purchases product

Attribution comparison:

• Last Click: Direct gets 100% credit, SEO gets 0%
• First Click: SEO gets 100% credit
• Linear: SEO gets 33%, Email gets 33%, Direct gets 33%
• Data-Driven: SEO might get 50% (discovery), Email 30% (nurture), Direct 20% (intent already existed)

Takeaway: In Last Click model, SEO looks worthless. In reality, it drove the discovery. Always use multi-touch attribution to properly value SEO.

What SEO metrics should I track?

Essential SEO metrics to track include: Organic traffic (sessions from search), Keyword rankings (positions for target keywords), Click-through rate from Google Search Console, Organic conversion rate (leads/sales from organic traffic), Backlink profile (referring domains, new/lost links), Core Web Vitals (LCP, FID, CLS), Indexed pages vs total pages, Average time on page and bounce rate, and Domain authority/rating.

Focus on metrics tied to business outcomes - traffic means nothing without conversions. Track trends over time, not just absolute numbers.

How do I set up Google Analytics 4 for SEO tracking?

Set up GA4 for SEO by:

1. Creating a GA4 property and adding the tracking code
2. Configuring key events (formerly conversions) like form submissions, purchases, downloads
3. Creating custom dimensions for landing page, user type, device category
4. Setting up enhanced measurement for scroll tracking and outbound clicks
5. Linking GA4 with Google Search Console
6. Creating custom reports for organic traffic analysis
7. Setting up traffic source segmentation to isolate organic search
8. Creating exploration reports for deep-dive analysis

GA4 focuses on events rather than pageviews.

What is statistical significance and why does it matter for SEO?

Statistical significance means your results are likely due to your changes, not random chance. In SEO, it prevents false conclusions from normal traffic fluctuations. For example, a 10% traffic increase might seem positive, but could be seasonal variation.

Test for significance by: running tests long enough (typically 4-8 weeks), accounting for seasonality, using larger sample sizes, applying statistical tests (chi-square test for CTR changes), and requiring 95%+ confidence level before declaring success. Don't make major decisions based on short-term fluctuations or small sample sizes.

How can I build an SEO dashboard?

Build an SEO dashboard using: Google Data Studio/Looker Studio (free, integrates with GA4 and Search Console), connect data sources (GA4, Google Search Console, your SEO tool's API), create key metrics widgets (organic traffic, rankings, CTR, conversions), add time-period comparisons (vs last month, vs last year), include segmentation by device, landing page, or query type, set up automated reports sent to stakeholders, and add filters for deeper analysis.

Popular tools: Data Studio, Tableau, Power BI, or all-in-one solutions like DashThis. Focus on actionable metrics, not vanity metrics.

What is cohort analysis and how do I use it for SEO?

Cohort analysis groups users by shared characteristics to track behavior over time. For SEO:

• Group users by acquisition date (when they first visited from organic search)
• Track how each cohort behaves over subsequent weeks/months (return rate, conversion rate, engagement)
• Compare cohorts to identify trends (are newer users engaging more?)
• Analyze by landing page cohorts (which entry pages lead to best long-term engagement)
• Segment by keyword theme or intent type

This reveals whether SEO improvements actually attract better-quality users who convert and return, not just more traffic.