Engagement · Gsc · Search · Ranking Changes
Overview

Analysis Overview

Analysis overview and configuration

Analysis TypeRanking Changes
CompanyMCP Analytics
ObjectiveCompare search performance between two time periods to identify ranking changes, click gains/losses, and new/lost pages
Analysis Date2026-03-02
Processing Idtest_1772497684
Total Observations318
ParameterValue_row
top_n30top_n
min_impressions10min_impressions
significance_level0.05significance_level
position_change_threshold2position_change_threshold
Interpretation

Purpose

This analysis compares search performance across two time periods to identify ranking shifts, click impact, and page portfolio changes for MCP Analytics. It establishes a baseline understanding of how the site's visibility and traffic have evolved, providing the foundation for understanding which content categories are gaining or losing search traction.

Key Findings

  • Matched Pages: 318 of 453 baseline pages tracked consistently—70% coverage with 94 pages lost and 96 new pages gained between periods
  • Winners vs. Losers: 32.4% of pages improved rankings (103 winners) while 23.3% declined (74 losers); 44.3% remained stable
  • Position Movement: Average improvement of -1.3 positions (negative = better ranking), but median change of -0.48 suggests most pages moved minimally
  • Click Paradox: Despite 2,656 additional impressions, total clicks dropped 65—indicating ranking gains aren't translating to click increases
  • Weak Correlation: Position-clicks correlation of -0.039 reveals ranking position has minimal influence on click behavior

Interpretation

The analysis reveals a mixed performance picture. While one-third of tracked pages improved their search rankings and impressions increased substantially, the overall click decline suggests either increased competition at better positions, lower click-through rates, or content relevance misal

Data preprocessing and column mapping

Initial Rows908
Final Rows318
Rows Removed590
Retention Rate35
Interpretation

Purpose

This section documents the data filtering applied before SEO performance analysis. The 65% data reduction indicates aggressive filtering criteria were applied to focus on meaningful comparisons, which directly impacts the reliability of conclusions about ranking changes and click performance across the matched page set.

Key Findings

  • Retention Rate: 35% (318 of 908 rows retained) - Indicates substantial filtering removed two-thirds of initial observations, likely pages below impression or engagement thresholds
  • Rows Removed: 590 observations excluded from analysis - Suggests minimum impression filter (10 impressions) and position threshold (2 positions) eliminated low-traffic pages
  • Matched Pages: 318 pages analyzed represent only the subset meeting quality criteria, excluding 94 baseline and 96 comparison unmatched pages from detailed analysis

Interpretation

The aggressive filtering prioritizes analysis reliability by excluding noise from low-traffic pages where position changes lack statistical significance. This aligns with the stated objective of analyzing meaningful SEO performance shifts. The 35% retention rate ensures that position changes (mean -1.3, median -0.48) and click patterns reflect substantive ranking movements rather than random fluctuation in marginal traffic sources.

Context

No train/test split is documented, suggesting this is exploratory analysis rather than predictive modeling. The filtering decisions directly explain why 590 pages appear in unmat

Executive Summary

Executive Summary

Executive summary and key takeaways

matched_pages
318
winners_count
103
losers_count
74
total_clicks_change
-65
overall_health
Mixed
MetricValue
Matched Pages318
Winners103 (32.4%)
Losers74 (23.3%)
Overall HealthMixed
Total Clicks Change-65
Bottom Line: Compared 453 baseline pages to 455 comparison pages, matching 318 pages for ranking change analysis. Overall SEO health: Mixed.

Key Findings:
• 32.4% of pages improved rankings (winners)
• 23.3% of pages declined in rankings (losers)
• 44.3% of pages remained stable
• Total traffic change: -65 clicks
• Position-clicks correlation: -0.039 (aligned with expected)

Recommendation: Investigate 74 declining pages to prevent further traffic loss. Document what worked for winners and apply those strategies to stable and losing pages.
Interpretation

EXECUTIVE SUMMARY

Purpose

This analysis compares SEO ranking performance across 318 matched pages between two periods to assess whether recent changes improved organic search visibility and traffic. The mixed health status indicates uneven performance requiring strategic intervention to reverse declining traffic trends.

Key Findings

  • Winners (32.4%, 103 pages): Achieved significant ranking improvements, with top performers moving up 7–45 positions; however, click gains were minimal (+2 total)
  • Losers (23.3%, 74 pages): Experienced ranking declines averaging +7.4 positions, contributing -7 clicks collectively
  • Stable Pages (44.3%, 141 pages): Largest segment but drove -60 clicks, indicating ranking stability masks traffic erosion
  • Overall Traffic Impact: -65 clicks despite 2,656 impression gains—suggesting ranking improvements aren't converting to clicks
  • Position-Click Correlation: -0.039 (near zero), confirming weak relationship between ranking position changes and click behavior

Interpretation

The portfolio shows a paradox: one-third of pages improved rankings substantially, yet total organic clicks declined. This disconnect suggests either that ranking improvements occurred in lower-traffic keywords, that click-through rates dropped despite better positions, or that impression gains came from lower-intent queries. The stable majority losing clicks indicates potential

Visualization

Position Change Distribution

Distribution of ranking changes across all pages

Interpretation

Purpose

This section visualizes the distribution of ranking position changes across all 318 matched pages, revealing whether the overall portfolio experienced improvements or declines. It serves as a foundational view of aggregate performance shifts, helping identify whether changes are concentrated in specific ranges or broadly distributed across the dataset.

Key Findings

  • Average Position Change: -1.30 positions - Indicates a slight net improvement in rankings across the portfolio, with negative values representing upward movement in search results
  • Median Position Change: -0.48 positions - Shows that the typical page experienced minimal ranking shift, suggesting stability despite some outliers
  • Distribution Range: -44.5 to +13.5 positions - Reveals extreme cases of both significant improvements (pages moving up 44+ positions) and notable declines (pages dropping 13+ positions)
  • Peak Concentration: Maximum bin count of 41 - Suggests clustering around a specific position change range, indicating most pages experienced similar magnitude shifts

Interpretation

The data reveals a portfolio with modest net ranking improvements offset by high variability. While the mean is negative (favorable), the median near zero indicates that most pages remained relatively stable, with performance gains concentrated among a smaller subset of high-movers. The wide range and right-skewed distribution (skew=-0.64) suggest that dramatic improvements are more common than dramatic declines, though the majority of

Visualization

Winners vs Losers Breakdown

Categorization of pages into winners, losers, and stable

Interpretation

Purpose

This section categorizes 318 matched pages into three performance tiers based on search position changes, using a 2.0-position threshold as the classification boundary. It provides a snapshot of overall SEO health by quantifying how many pages improved, declined, or remained relatively stable during the comparison period—essential context for understanding whether the domain experienced net positive or negative momentum.

Key Findings

  • Winners (103 pages, 32.4%): Pages improved by 2+ positions with an average gain of -7.3 positions and +2 total clicks, indicating meaningful ranking improvements despite minimal click volume changes.
  • Losers (74 pages, 23.3%): Pages declined by 2+ positions with an average loss of +4.88 positions and -7 total clicks, representing the smallest segment but showing measurable ranking deterioration.
  • Stable (141 pages, 44.3%): Nearly half of matched pages shifted less than 2 positions, yet this group absorbed -60 clicks—the largest click loss despite positional stability, suggesting ranking volatility or SERP feature changes.

Interpretation

The distribution reveals a mixed SEO performance picture: while winners outnumber losers (103 vs. 74), the stable majority (44.3%) masks concerning click losses. Winners' modest position improvements (+7.3 average

Visualization

Top Movers

Top ranking winners and losers with largest position changes

Interpretation

Purpose

This section isolates the 60 pages with the most dramatic ranking shifts to enable targeted investigation of outlier performance. By examining extreme movers—both winners and losers—you can identify what factors drive significant ranking changes and understand whether position improvements translate to meaningful click gains or losses.

Key Findings

  • Largest Winner: -44.89 position change (improved from position 52.6 to 7.71) with 0 click change, indicating strong ranking recovery without immediate click impact
  • Largest Loser: +13.9 position change (declined from position 11.33 to 25.23) with 0 click change, showing ranking deterioration without proportional click loss
  • Distribution Asymmetry: Winners show larger absolute position changes (mean -16.32) compared to losers (mean +7.41), suggesting ranking improvements are more dramatic than declines
  • Click-Position Disconnect: Mean clicks_change of -0.05 across top movers reveals weak correlation between position shifts and click volume, even for extreme movers

Interpretation

The top movers dataset reveals that ranking position changes don't automatically drive click changes. Pages improving 40+ positions show zero click gains, while pages declining 10+ positions maintain stable clicks. This suggests either a lag effect in click response, or that position alone doesn't guarantee traffic conversion

Visualization

Traffic Impact Analysis

Relationship between position changes and traffic impact

Interpretation

Purpose

This section examines whether ranking improvements translate into traffic gains—a critical measure of SEO effectiveness. The correlation analysis reveals the strength of the relationship between position changes and click changes, helping determine if ranking movements are actually driving user engagement or if other factors are influencing traffic.

Key Findings

  • Correlation Coefficient (-0.039): Extremely weak negative correlation indicates ranking changes have minimal predictive power over click changes. Pages moving up in rankings are not consistently gaining clicks.
  • Total Clicks Change (-65): Overall traffic declined by 65 clicks despite 103 pages improving rankings, suggesting ranking improvements alone did not drive traffic recovery.
  • Position Change Distribution: Mean position improvement of -1.3 positions across matched pages, yet median clicks change is 0, indicating most pages saw no traffic impact from ranking shifts.

Interpretation

The near-zero correlation reveals a disconnect between ranking performance and traffic outcomes. While 32.4% of pages improved rankings (winners), the aggregate click loss suggests external factors—content relevance, search intent misalignment, or competitive dynamics—may be limiting traffic gains from better positions. This pattern indicates ranking improvements alone are insufficient to explain traffic changes in this dataset.

Context

This analysis assumes position and clicks are the primary SEO metrics; other factors like impressions (which increased by 2,656) or conversion quality are not captured here.

Data Table

Top Winners Detail

Detailed metrics for top ranking winners

page_urlpage_fullpositionposition.1position_changeclicks_changeimpressions_changeclicksclicks.1
what-we-learned-analyzing-shopify-stores-with-p...https://mcpanalytics.ai/blogs/what-we-learned-analyzing-shopify-stores-with-product-price-elasticity-analysis.html52.67.7-44.90-800
whitepaper-multi-echelon-optimizationhttps://mcpanalytics.ai/whitepapers/whitepaper-multi-echelon-optimization56.212.6-43.60-14400
t-test-guidehttps://mcpanalytics.ai/articles/t-test-guide65.526.2-39.30-7700
blog-shopify-product-bundle-affinity-analysishttps://mcpanalytics.ai/blogs/blog-shopify-product-bundle-affinity-analysis37.46.7-30.70-5100
ab-testinghttps://mcpanalytics.ai/ab-testing33.57.2-26.2-1-10410
blog-squarespace-shipping-cost-efficiencyhttps://mcpanalytics.ai/blogs/blog-squarespace-shipping-cost-efficiency3610.9-25.10-2100
whitepaper-fee-breakdownhttps://mcpanalytics.ai/whitepapers/whitepaper-fee-breakdown38.615.7-22.90-3400
hybrid-recommender-system-practical-guide-for-d...https://mcpanalytics.ai/articles/hybrid-recommender-system-practical-guide-for-data-driven-decisions51.328.5-22.901100
linear-discriminant-analysis-lda-practical-guid...https://mcpanalytics.ai/articles/linear-discriminant-analysis-lda-practical-guide-for-data-driven-decisions.html31.413.6-17.80-311
cash-flow-forecasting-practical-guide-for-data-...https://mcpanalytics.ai/articles/cash-flow-forecasting-practical-guide-for-data-driven-decisions24.18.7-15.40-900
whitepaper-propensity-score-matching.htmlhttps://mcpanalytics.ai/whitepapers/whitepaper-propensity-score-matching.html25.711.2-14.50-400
blog-woocommerce-order-value-segmentation-analysishttps://mcpanalytics.ai/blogs/blog-woocommerce-order-value-segmentation-analysis216.7-14.307600
blog-shopify-average-order-value-analysishttps://mcpanalytics.ai/blogs/blog-shopify-average-order-value-analysis22.18.2-13.90-2700
how-to-use-inventory-status-in-shopify-step-by-...https://mcpanalytics.ai/tutorials/how-to-use-inventory-status-in-shopify-step-by-step-tutorial.html22.89-13.70300
blog-ebay-ebay-orders-status-trackinghttps://mcpanalytics.ai/blogs/blog-ebay-ebay-orders-status-tracking22.710.4-12.30900
the-woocommerce-mistake-thats-costing-you-money...https://mcpanalytics.ai/blogs/the-woocommerce-mistake-thats-costing-you-money-and-how-to-fix-it17.35.4-11.90300
whitepaper-synthetic-controlhttps://mcpanalytics.ai/whitepapers/whitepaper-synthetic-control15.15.2-9.901400
ab-testing-statistical-significancehttps://mcpanalytics.ai/articles/ab-testing-statistical-significance25.415.7-9.70-12200
how-to-use-failed-payment-recovery-analysis-in-...https://mcpanalytics.ai/tutorials/how-to-use-failed-payment-recovery-analysis-in-stripe-step-by-step-tutorial.html18.28.9-9.30-7700
how-to-use-discount-effectiveness-in-etsy-step-...https://mcpanalytics.ai/tutorials/how-to-use-discount-effectiveness-in-etsy-step-by-step-tutorial17.68.3-9.3117801
whitepaper-factor-analysis.htmlhttps://mcpanalytics.ai/whitepapers/whitepaper-factor-analysis.html2212.8-9.2013400
customer-lifetime-value-ltv-practical-guide-for...https://mcpanalytics.ai/articles/customer-lifetime-value-ltv-practical-guide-for-data-driven-decisions17.58.6-8.90-4400
cox-proportional-hazards-practical-guide-for-da...https://mcpanalytics.ai/articles/cox-proportional-hazards-practical-guide-for-data-driven-decisions.html30.221.4-8.8054300
gaussian-mixture-models-practical-guide-for-dat...https://mcpanalytics.ai/articles/gaussian-mixture-models-practical-guide-for-data-driven-decisions.html24.215.7-8.5212902
analytics__economics__elasticity__pricehttps://mcpanalytics.ai/services/analytics__economics__elasticity__price156.5-8.5-11510
logistic-classification-practical-guide-for-dat...https://mcpanalytics.ai/articles/logistic-classification-practical-guide-for-data-driven-decisions.html18.910.8-8.1-15010
commerce__square__customers__repeat_customer_an...https://mcpanalytics.ai/analysis/reports/commerce__square__customers__repeat_customer_analysis23.515.6-7.90-300
how-to-use-discount-effectiveness-in-etsy-step-...https://mcpanalytics.ai/tutorials/how-to-use-discount-effectiveness-in-etsy-step-by-step-tutorial.html16.99-7.8-14810
whitepaper-fishers-exact.htmlhttps://mcpanalytics.ai/whitepapers/whitepaper-fishers-exact.html20.313.1-7.2115712
whitepaper-market-basket.htmlhttps://mcpanalytics.ai/whitepapers/whitepaper-market-basket.html13.46.3-7.101700
Interpretation

Purpose

This section isolates the 30 pages with the largest ranking improvements to identify replicable SEO success patterns. Winners demonstrate which content strategies, updates, or optimizations drove measurable position gains—providing a blueprint for applying similar tactics across the broader site portfolio.

Key Findings

  • Average Position Improvement: -16.32 positions (mean) — Winners moved from position 27.88 to 11.55, representing substantial first-page visibility gains
  • Position Range: -44.9 to -7.1 positions — Extreme variability shows both dramatic recoveries (e.g., position 52.6→7.7) and modest gains
  • Click Impact Disconnect: Mean clicks_change = 0, median = 0 — Most winners gained ranking without proportional click increases, suggesting position improvements haven't yet translated to traffic
  • Impressions Volatility: Mean change = +21.97, but SD = 124.31 — Highly inconsistent impression changes indicate winners don't follow a uniform pattern

Interpretation

Winners achieved significant ranking improvements, but the weak correlation between position gains and click increases suggests either: (1) pages are still below optimal click-through positions, (2) query intent misalignment persists, or (3) ranking improvements are recent and haven't fully matured. The diversity in baseline positions (13

Data Table

Top Losers Detail

Detailed metrics for top ranking losers

page_urlpage_fullpositionposition.1position_changeclicks_changeimpressions_changeclicksclicks.1
whitepaper-spectral-clustering.htmlhttps://mcpanalytics.ai/whitepapers/whitepaper-spectral-clustering.html11.325.213.901700
articleshttps://mcpanalytics.ai/articles/10.422.311.90800
support-vector-machine-svm-practical-guide-for-...https://www.mcpanalytics.ai/articles/support-vector-machine-svm-practical-guide-for-data-driven-decisions.html8.82011.2-2-47120
whitepaper-spectral-clusteringhttps://mcpanalytics.ai/whitepapers/whitepaper-spectral-clustering11.121.810.704300
whitepaper-neural-networkshttps://mcpanalytics.ai/whitepapers/whitepaper-neural-networks7.51810.50-1000
porter-five-forces-analysis-practical-guide-for...https://mcpanalytics.ai/articles/porter-five-forces-analysis-practical-guide-for-data-driven-decisions6.316.410.11-1601
whitepaper-group-lassohttps://mcpanalytics.ai/whitepapers/whitepaper-group-lasso4.214.19.90-16800
whitepaper-ldahttps://mcpanalytics.ai/whitepapers/whitepaper-lda6.816.29.40800
what-we-learned-analyzing-etsy-stores-with-prod...https://mcpanalytics.ai/blogs/what-we-learned-analyzing-etsy-stores-with-product-mix.html816.88.80-500
whitepaper-chi-square.htmlhttps://mcpanalytics.ai/whitepapers/whitepaper-chi-square.html3.912.18.1016900
how-to-use-geographic-sales-analysis-in-woocomm...https://mcpanalytics.ai/tutorials/how-to-use-geographic-sales-analysis-in-woocommerce-step-by-step-tutorial.html10.518.37.8-1-1910
whitepaper-naive-bayeshttps://mcpanalytics.ai/whitepapers/whitepaper-naive-bayes5.913.67.70000
whitepaper-vehicle-routinghttps://mcpanalytics.ai/whitepapers/whitepaper-vehicle-routing5.7137.20-1400
support-vector-machine-svm-practical-guide-for-...https://mcpanalytics.ai/articles/support-vector-machine-svm-practical-guide-for-data-driven-decisions916.17.124602
k-means-clustering-practical-guide-for-data-dri...https://mcpanalytics.ai/articles/k-means-clustering-practical-guide-for-data-driven-decisions8.315.47.11-10801
anova-practical-guide-for-data-driven-decisionshttps://mcpanalytics.ai/articles/anova-practical-guide-for-data-driven-decisions7.514.570-19700
whitepaper-voting-ensemblehttps://mcpanalytics.ai/whitepapers/whitepaper-voting-ensemble7.313.46.104000
whitepaper-fishers-exacthttps://mcpanalytics.ai/whitepapers/whitepaper-fishers-exact13195.9229313
xgboost-practical-guide-for-data-driven-decisio...https://mcpanalytics.ai/articles/xgboost-practical-guide-for-data-driven-decisions.html14.620.55.9-236720
what-we-learned-analyzing-square-stores-with-ho...https://mcpanalytics.ai/blogs/what-we-learned-analyzing-square-stores-with-hourly-performance-analysis5.811.55.70-2800
whitepaper-glmhttps://mcpanalytics.ai/whitepapers/whitepaper-glm5.711.35.7-3-1541
whitepaper-revenue-analysishttps://mcpanalytics.ai/whitepapers/whitepaper-revenue-analysis5.811.35.50900
difference-in-differences-practical-guide-for-d...https://mcpanalytics.ai/articles/difference-in-differences-practical-guide-for-data-driven-decisions.html4.810.15.30-17200
whitepaper-propensity-score-matchinghttps://mcpanalytics.ai/whitepapers/whitepaper-propensity-score-matching1015.35.207700
whitepaper-vehicle-routing.htmlhttps://mcpanalytics.ai/whitepapers/whitepaper-vehicle-routing.html7.212.35.125702
whitepaper-feature-importancehttps://mcpanalytics.ai/whitepapers/whitepaper-feature-importance7.812.85-31230
how-to-use-item-modifier-analysis-in-square-ste...https://mcpanalytics.ai/tutorials/how-to-use-item-modifier-analysis-in-square-step-by-step-tutorial9.714.850-8000
whitepaper-pcahttps://mcpanalytics.ai/whitepapers/whitepaper-pca10.415.24.70-1100
whitepaper-pca.htmlhttps://mcpanalytics.ai/whitepapers/whitepaper-pca.html9.714.14.40-700
whitepaper-kaplan-meierhttps://mcpanalytics.ai/whitepapers/whitepaper-kaplan-meier6.811.14.40-4500
Interpretation

Purpose

This section isolates the 30 pages with the largest ranking declines to enable targeted recovery efforts. By examining pages that lost the most search visibility, the analysis identifies which content requires immediate investigation for technical issues, competitive displacement, or algorithmic impact. This complements the overall analysis by highlighting underperformance patterns that offset gains from winning pages.

Key Findings

  • Average Position Decline: 7.41 positions—a substantial drop indicating these pages have fallen significantly in search results
  • Baseline Position: Mean of 8.13 (strong starting rank) declining to 15.55, showing previously competitive content losing visibility
  • Click Impact: Minimal correlation (mean -0.1 clicks lost)—most losers experienced negligible traffic loss despite ranking drops
  • Impression Volatility: High variance (SD=144.26) with mixed changes, suggesting inconsistent visibility patterns across this cohort
  • Content Type: Whitepapers and educational articles dominate, indicating technical/reference content may be underperforming

Interpretation

The losers represent a paradox: significant ranking deterioration without proportional click losses. This suggests these pages ranked for competitive, lower-intent keywords where position matters less, or they occupy lower-CTR positions (10+). The concentration of whitepapers indicates potential content freshness or topical authority issues in technical domains

Want to run this analysis on your own data? Upload CSV — Free Analysis See Pricing